Probabilistic Logical Models (PLMs) combine the powerful formalisms of probability theory and first-order logic to handle uncertainty in large, complex problems. While PLMs provide a very effective learning paradigm under the umbrella of Statistical Relational Learning (SRL) methods, tractable inference is a significant problem in these models. Earlier approaches focused on grounding the model to a propositional network to use existing inference algorithms. Other popular techniques include sampling and lifted inference, with a lot of interest in the latter recently.

In this talk, I will present three different approaches to accelerate inference in PLMs. First, a preprocessing method for Markov Logic Networks that makes exact grounded inference tractable; second, an approximate inference method called `counting belief propagation' that performs belief propagation on compressed factor graphs; and finally, an `anytime'inference algorithm that returns a bound over the marginal distribution of the query variable. I will present experimental results to demonstrate the usefulness of these three distinct, yet related, inference methodologies.

Biography

Sriraam Natarajan is currently a Post-Doctoral Research Associate at the Department of Computer Science at University of Wisconsin-Madison. He graduated with his PhD from Oregon State University working with Dr. Prasad Tadepalli. His research interests lie in the field of Artificial Intelligence, with emphasis on Machine Learning, Statistical Relational Learning, Reinforcement Learning, Graphical Models and Bio-Medical Applications.

Intelligent scissors is an interactive image segmentation tool that allows a user to select piece-wise globally optimal contour segments (based on an optimal path search in a graph) that correspond to a desired object boundary. Our current work uses tobogganing to raise the granularity of the image primitive above the pixel level, producing a region-based basic processing unit that is object-centered rather than device-dependent. The resulting region-based elements form the basis several improvements to intelligent scissors that reduce (often greatly) the human time and effort needed for object selection while simultaneously increasing the accuracy of boundary definition. Included in these improvements is an edge confidence measure that allows for improved cursor snapping, automatic path extension, and the ability to select certain objects simply by snapping to them. Also, the tobogganed-based regions provide for an edge model that computes subpixel boundary position, noise-independent edge blur adjustment, and automatic alpha matte generation and color seperation of boundary transition pixels--thereby improving the quality of certain image editing operations such as cut and paste.

Biography

Dr. Eric Mortensen joined the Computer Science Department at Oregon State University in Aug. 2001 after completing a Ph.D. in Computer Science at Brigham Young University. During his graduate work, Dr.Mortensen developed and extended a "cutting edge" user-guided image segmentation techniques called intelligent scissors. In addition to image and video segmenation and editing, Dr. Mortensen's other research interests include image-based modeling and interactive vision and graphics techniques.

Personal Digital Assistants (PDAs), cell phones, wireless networks and instant messaging have opened new opportunities for collaborative ebusiness and enterprise computing. Software agents, personalization techniques, user interface technologies and integration architectures are effective building blocks for engineering scalable middleware solutions or "web-ware". It is now possible to envisage bolt-on web-ware solutions that incrementally extend existing applications for collaborating end-users and enterprise workers. This seminar provides an overview of an on-going web-ware project at The Wise Net Inc. and directly related R&D at OSU. The development and R&D teams are exploring the application of Jini/Javaspaces, SOAP, XML, constraint processing and security technologies to build collaborative middleware for the real estate industry. Technical challenges from the perspective of the software engineer will be presented. Current OSU R&D projects and problem areas needing to be addressed are outlined.

Biography

Kal Toth is an Associate Professor at Oregon State University in the Department of Computer Science. He has over 25 years of industry experience in the fields of software engineering, electronic commerce, project management, distributed information networks and information security. He is actively conducting research in the field of ebusiness investigating wireless and wire-line personalization, security and interoperability problems using intelligent software agents and web technologies.

Visualization research is the science and art of designing, implementing and evaluating methods for effectively communicating information through images. Fundamental questions include: How can we determine how best to portray a large, complicated set of data so that its essential features can be accurately and intuitively understood? How can we best measure the success of our efforts? Where should we look to gain insight into the science behind the art of effective visual representation? In this talk I will address these and other visualization research issues in the context of some of my recent work in designing effective textures for 3D shape representation and 2D multivariate flow visualization. In particular, I will describe our efforts to unite inspiration from art with insights from fundamental findings in human visual perception to define the characteristics of texture patterns that can be effectively used for visualization purposes such as facilitating accurate shape perception and the integrated comprehension of multivariate distributions. As part of this discussion, I will present the findings of our recent observer experiments intended to evaluate the effects of texture orientation, and other characteristics, on shape judgment accuracy. Time permitting, I will also describe some of our recent work in developing methods to more accurately estimate the continuous geometric properties of smooth surfaces approximated by arbirary triangle meshes, and in efficiently synthesizing shape-conforming textures.

CREEDA is an integrated Web-based application for planning agricultural activities. This application incorporates Web versions of such applications as RUSLE (Revised Universal Soil-Loss Equation), STMDL (Sediment Total Maximum Daily Load), WebPST (Web-Based Pesticide Screening Tool), SISL (Surface Irrigation Soil Loss), and SCI (Soil Conditioning Index). Web-based map interfaces are an important feature of these applications. Each application uses a map generated by ArcIMS as the primary user interface. With this map interface, the user can insert, query, update, and delete map features and the information associated with them. A relational database and an ArcSDE server in combination manage the geospatial data. The ArcIMS Internet map server generates maps to be displayed on a Web browser by using the geospatial data provided by the ArcSDE server. The Web pages, including those that display maps, are generated by the server-side scripts written in ASP.NET.

Biography

Measured impedances of mm-wave components in the B.S. project. Designed and implemented a digital control system for a rate-gyroscope in the M.S. project. Enagaged in design and implementation of computer-control systems for fossil-fuel and nuclear power plants for six years. Obtained a Ph. D. degree by working on concurrency and recovery schemes for distributed database systems. Created many sample programs as active object systems. Currently working on Web-based applications that use databases and GIS interfaces. A licensed first-class radio engineer (Japan) and a licensed fisherman (Oregon).

Tools and environments to enable end users to "program" are becoming increasingly popular. The best known such environment is the spreadsheet, and the way users program in this type of environment is by providing formulas. In this talk, we will consider what happens when we add to end-user programming environments consideration of elements of the software engineering lifecycle beyond coding. Doing so seems necessary, because there is ample evidence that end users' programs are no more reliable than those written by professional software engineers. My colleagues and I have been developing a holistic approach to software engineering for end users. It incorporates support for testing, finding bugs, and specification, in an incremental manner integrated in a fine-grained way with the programming environment. The software engineering knowledge needed is in the system, and the user is not expected to develop expertise at software engineering; instead, the strategy is for the system to provide guidance to the user. In the talk, I will focus primarily on how testing and assertions are supported as part of this holistic approach. A number of papers on this topic are available at http://www.cs.orst.edu/~burnett/reprints.html

Biography

Margaret Burnett has been involved in visual programming language research for many years, and most recently her interests have turned especially to end-user programming and end-user software engineering. She is the principal architect of the Forms/3, a spreadsheet-like research language for exploring the boundaries of the spreadsheet paradigm, and of the FAR multi-paradigm programming language for end users. Burnett was recently honored with Oregon State University's Elizabeth P. Ritchie Distinguished Professor Award. She is also a past recipient of the National Science Foundation's Young Investigator Award. She has been a member of the Program Committees for the IEEE Visual Language Symposium, ACM Conference on Programming Language Design and Implementation, and the ACM Conference on Functional Programming, and several others. She has also held various offices in conference committees of the IEEE Visual Language Symposium, ACM Multimedia, and ACM/IEEE International Conference on Software Engineering. For more information about her research and/or papers, see http://www.cs.orst.edu/~burnett/

At the heart of the success of the Internet is its congestion control behavior. TCP is the dominant congestion control protocol in the Internet, and it determines the bandwidth allocation among TCP flows. In recent years, multimedia applications have become increasingly popular in the Internet. TCP, although powerful and effective, is not sufficient to satisfy multimedia application well. Therefore, many congestion control protocols have been proposed recently particularly for streaming media in the Internet. To ensure that flows using a new protocol share bandwidth fairly with TCP flows, we have to predict the bandwidth share ratio among them before deploying new protocols. The prediction can be done in a static way that assumes a constant congestion signal for all traffic. It can also be done in a dynamical way that associates each individual flow's congestion perception to its rate behavior. We choose a dynamic approach and model the bandwidth share among competing traffic with a state-space model. We use the model to describe the stability of bandwidth competitions, which is characterized as convergence to a dynamically oscillating limit cycle in the state space. Real-world experiments confirmed to us that the dynamic modeling produces results that closely match real measurements, while some statically derived share ratios sometimes do not.

Biography

Kang Li is a Postdoc in the System Software Laboratory in the Computer Science and Engineering Department at Oregon Graudate Institute (OGI) in Beaverton, Oregon. He received his Ph.D in Computer Science and Engineering from OGI in October, 2002. His research interests are in the area of computer networks, in particular, congestion control and network measurements.

As Open Source software has matured, so have the tools that enable their development, production, distribution, and use of feedback management. Web-based Portal Communities issued under the General Public License (GPL) that support Open Source software development have sprung up virtually around the world. One significant tools source has been the VASoftware SourceForge.net application. The application integrates Task Management, Bug Tracking, Support Request handling, file distribution, database I/O, and source configuration management (SCM) into a common Web-enabled platform. SourceForge success has been ensured through its use by more than 500,000 software developers working on more than 50,000 projects. Many SourceForge-based communities have been formed beyond the VASoftware sponsored portal. Some communities are open to the public at large. Others, in fact a significant number, have been brought behind corporate firewalls and provide the basic tools needed to efficiently produce software in a closed setting. Expanding SourceForge deployments one step further, Tektronix has made a donation of a fully functioning SourceForge Community to Oregon State University. It is called the Oregon Software Technology Exchange (OSTE). The power and potential of OSTE will be presented and discussed. The impact that OSTE provides by integrating Web-based software development tools will be illustrated.

Biography

Christopher Perez currently program manages initiatives which span Tektronix, Inc. engineering, manufacturing, corporate IT, and customer service functions. He is helping to solve challenges involving globally distributed software design centers, timezones, countries, and cross cultural information exchanges are addressed. He is helping investigate solutions to global software engineering project management challenges which include technical information, software configuration management, defect tracking, production release, and intellectual property distribution, management, and historical archive creation. Christopher's 20 year employment history includes providing products covered under the Open Source Movement, X terminal and Test and Measurement product development, WAN and LAN RF client server technology deployments, and aerospace software design and implementation. The Tektronix Web site is www.tektronix.com

Software as a technology is fairly well understood, but software as a business is not. The relation between the two is so unique, challenging and important that some business schools and computer science departments have begun to address it specifically. This is timely, for while the majority of software professionals continue to work in IT departments of companies whose business is not software, these jobs are increasingly eroded by commercial software and outsourcing. Until it fully embraces the issues of software commercialization, software development cannot achieve relative security as a career or true engineering status as a discipline, nor can it fulfill its inherent potential to build significant personal and economic wealth. Effective strategic planning for today's high-volume software markets requires an understanding of how software market structure evolves, and how these markets move toward value. This requires a new paradigm, or way of seeing software value. The conventional view that software value derives mainly from internalities, such as features or quality, is dangerously blind to the critical externalities that come to dominate actual perceived value in higher volume software products. Successful high-volume software products are designed for total value and are managed as an evolving release portfolio with a compelling theme and trajectory that recruits and orients customer demand. Certain product strategies can reorient demand away from competitors' trajectories, dramatically changing the direction of both the technology and the market. This talk, targeted to both computer science and business professionals, gives a brief overview of modern value driven software commercialization issues and strategies.

Biography

Frank Hall has more than 25 years experience in software development, management and research. He is president of EntreDigm Consulting LLC, a corporate and business development consultancy for software and Internet companies (www.entredigm.com). He is a co-founder and board director of Infinity Softworks, the global market leader in software calculators for handheld computers (www.infinitysw.com), and a former board director of Camo, an international statistical software and services corporation (www.camo.com). He serves as a business development executive for ProWorks, maker of data visualization software (www.proworks.com), and has similarly served Infinity Softworks, Camo and PrintQuick, maker of Internet print delivery software (www.printquick.com). Frank Hall is an adjunct faculty member of OHSU's OGI School of Science and Engineering, where he has taught a graduate class in software commercialization for OHSU and PSU (www.ogi.edu/MST/classes/MST531). He is a board director of the Software Association of Oregon, and serves as president of its Corvallis chapter. Previously he served Hewlett-Packard for 12 years as an R&D project manager developing open systems software, including the Motif user interface toolkit. His research in the mid-1990s into software business success factors led him to leave HP and launch his consulting firm. He has published in IEEE Software and the HP Journal. He completed all but thesis toward an MS in Computer Science at OSU, and holds an MS degree in cultural anthropology from the University of Texas at Austin, and a BA in mathematics from Florida State University. He lives in Corvallis with his wife and two children.

If the level of adoption of Software Engineering Best Practice is to be increased in industry, then it must be taught effectively in the university. Many Computer Science students, however, either view software process as intellectually shallow or are averse to the oppressive discipline which they perceive to be required to follow it. We have devised a method, Live-Thru Case Histories, for motivating students to recognize the necessity of Software Engineering Best Practice, and to learn how to use it, by shocking them into the realization that without it they are likely to fail, not in their course work, but in real-world software development projects. The method has been used for three years as part of a required Senior Project/Software Engineering course, at Stevens Institute; it was recently used in Barry Boehm's graduate Introduction to Software Engineering at USC, and has been used at a number of other universities to good effect. Its effectiveness is being assessed through the use of a number of instruments which we have developed, including an Attitude Toward Software Engineering (ATSE) survey. ATSE has been validated through administration and focus groups at meetings of NJSPIN (North Jersey Software Process Improvement Network), LASPIN, Southern California SPIN, Xerox Corporation's Software Engineering Process Group, and DoD's Software Technology Conference. We discuss the Live-Thru method, the use of ATSE in assessing its effectiveness, and the more general use of ATSE in outcomes assessment of Software Engineering courses, degree programs in Computer Science and Software Engineering and of software process improvement efforts in industry.

Biography

Prof. Klappholz has a BS in mathematics and linguistics from MIT, and an MSEE and PhD in Computer and Information Science from the University of Pennsylvania. He has taught at Columbia University, Polytechnic University, and Stevens Institute of Technology. His technology research areas have included parallel computer architecture, compilation techniques for parallel architectures, and general compiler optimization techniques. He has recently become interested in software development process. His most recent research interests include techniques for measuring attitude toward, knowledge of, and ability to apply, software development process/best practice, pedagogic techniques for teaching process/best practice, empirical software engineering studies, especially within the classroom, and the extension of static, point solution cost/schedule/quality estimation tools to automated tools for continuous estimation and (feedback-based) software project management. He has recently completed a sabbatical at USC, where he worked with Barry Boehm and Dan Port on various software development research and pedagogy issues.

Camera calibration is one of the classical problems in Computer Vision. It consists in estimating the internal geometry of the camera, namely the image formation process through its optical system, as well as the external geometry of the camera, i.e., its location and orientation in space. The pinhole camera is a convenient abstraction for modeling any optical acquisition device: 3-D points are mapped into 2-D image points through a simple perspective projection. However, unlike the ideal pinhole camera, any real camera is equipped with lenses which inevitably introduce image distortion. The estimation of the parameters describing such distortion and the estimation of the actual focal length of the camera provide the complete characterization of the internal geometry of the camera. The knowledge of the external geometry is not necessary in many applications but it has to be obtained anyway since its parameters are intertwined with those of the internal geometry. In this colloquium, I will present a new and very efficient method for the complete calibration of a digital camera.

Biography

Dr. Luca Lucchese received the M.S. degree and the Ph.D. in electrical engineering from the University of Padua, Italy, in 1993 and 1997, respectively. From 1997 to 2002, he was with the Department of Electrical and Computer Engineering of the University of California, Santa Barbara, first a post-doctoral researcher and then as a visiting assistant professor. Since March 2002, he has been with the Department of Electrical and Computer Engineering at Oregon State University as an assistant professor. He teaches courses of digital signal processing and image processing. His research interests include motion analysis and estimation, image registration and mosaicking, three-dimensional imaging, and color image processing. Dr. Lucchese is an Associate Editor of the IEEE Transactions on Image Processing.

The TaskTracer Project is a newly initiated computing research project at Oregon State University that is investigating the possibilities of a desktop software system that will track in detail how knowledge workers complete tasks, and intelligently leverage that information to increase efficiency and productivity. The goal is to create a system with four capabilities: more efficient task-interruption recovery, individual knowledge management, workgroup knowledge management, and within-workgroup workflow detection and analysis. The proposed system will operate in the Microsoft Windows environment, tracking most interactions with desktop applications as well as tracking phone calls. A central challenge involves “unweaving” the threads of events that are affiliated with different tasks that are being performed concurrently. Our approach will combine creative user interfaces and machine learning to perform this unweaving. Other challenges include identifying “significant” events, summarizing tasks or subtasks, predicting likely steps and resources, and detecting cross-user workflow. The TaskTracer project at Oregon State is part of the Management of Knowledge-Intensive Dynamic Systems (MKIDS) initiative funded jointly by the National Science Foundation and the Intelligence Community.

Biography

Jon Herlocker (BS Lewis & Clark College 1994; PhD Minnesota 2000) is Assistant Professor of Computer Science at Oregon State University. Dr. Herlocker's research work centers on integrating intelligence and usability into information systems. At the University of Minnesota, he was part of a group of researchers that developed a collaborative filtering-based recommender system which was licensed to Net Perceptions Inc, and was the lead designer and developer for the MovieLens web-based movie recommender (www.movielens.org). He has published a number of scientific papers on algorithms and systems for collaborative filtering, and has been awarded the prestigious NSF Faculty Early Career Development (CAREER) award for his work in that area. Dr. Herlocker is a member of the ACM and ACM SIGCHI, the IEEE and the Computer Society, and is the faculty advisor the local student chapter of the ACM. http://cs.oregonstate.edu/~herlock/

The schema for a relational database contains key information about the structure of the information it contains. Exploitation of this information can substantially enhance the power and ease-of-use of machine-learning and data-mining systems. In this talk I will review recent and ongoing research in developing a tractable language for specifying derived variables in probabilistic model discovery from relational data sources.

Biography

Dr. D'Ambrosio is an Associate Professor in Computer Science at Oregon State University. He is also founder of CleverSet, Inc. His research focuses on representation, inference, and discovery of relational probabilistic models, especially for such tasks as situation assessment and process modeling (e.g., web user behavior, west-nile virus spread)

http://www.cs.orst.edu/~dambrosi

The People and Practices Research Lab at Intel consists of about a dozen social scientists, designers, and engineers. We study real people across the globe in their natural work and life environments in order to imagine new uses for computing power, to identify unmet needs and desires, and to understand barriers to technology adoption. In this talk, I will introduce the methods and perspective of the group, using examples from recent work in US living rooms, Korean homes and "PC rooms", and Chinese workplaces.

Biography

Scott Mainwaring joined Intel's People and Practices Research Lab in 2000. His research interests include: community, trust, and the social use of technologies; new technologies for homes and families; and the influence of geography and culture on technology adoption and use. Prior to joining Intel, he was a researcher at Interval Research Corp. in Palo Alto, CA for six years, conducting ethnographic fieldwork and prototyping studies in order to better understand the real and potential roles of technology in everyday life. In earlier lives, he could be found developing Unix applications, administering an undergraduate cognitive science program, or studying the formation of spatial mental models in college sophomores. Scott holds an A.B. in computer science from Harvard University and a Ph.D. in cognitive psychology from Stanford University. (See http://www.intel.com/research/people/mainwaring_s.htm)

Most cost-effective digital cameras use a single image sensor, applying alternating patterns of red, green, and blue color filters to each pixel location. A way to reconstruct a full three-color representation of color images by estimating the missing pixel components in each color plane is called a demosaicing algorithm. This paper presents three inherent problems often associated with demosaicing algorithms that incorporate directional interpolation: misguidance color artifacts, interpolation color artifacts, and aliasing. The level of misguidance color artifacts present in two images can be compared using metric neighborhood modeling. The proposed demosaicing algorithm estimates missing pixels by interpolating in the direction with fewer color artifacts. The aliasing problem is addressed by applying filterbank techniques to directional interpolation. The interpolation artifacts are reduced using a nonlinear iterative procedure. Experimental results using digital images confirm the effectiveness of this approach.

Biography

Keigo Hirakawa received the B.S. degree in electrical engineering from Princeton University, Princeton, NJ, in 2000. He is currently pursuing the MS/Ph.D. degree at Cornell University, Ithaca, NY. His research interests include image modeling, color representation, multi-rate systems, and image interpolation. He also pursues a professional career as a jazz pianist.

In a balanced code each code word contains equal number of 1's and 0's. These codes find applications in many areas; they can be used to detect unidirectional errors in VLSI systems, to design fault tolerant and fail-safe sequential circuits, to achieve data integrity in write-once memories, as line codes in fiber optic data transmission systems, as modulation codes in optical and magnetic storage systems, for high speed VLSI design, etc. In coding theory, efficient design of encoding and decoding of balanced codes (i.e. converting data words to balanced codes and from balanced words to data words) has been an open research problem for many years. In 1986, Knuth has given some simple design schemes for these codes. Since then we have given many improved design methods. In this talk, after describing some of the applications, some design methods will be described.

Biography

http://cs.oregonstate.edu/~bose/cv.html

Walden's Paths enables the specification and viewing of presentations formed out of materials gathered from throughout the World-Wide Web. This is accomplished by layering a metastructure--in this case a linear path--on top of already existing materials. The organization of materials into paths allows (indeed requires) the inclusion of annotation in order to provide contextualization to the reader. Ownership of Web materials is administratively decentralized, and this raises challenges for Path maintenance; essentially, a goal of Walden's Paths is to build a stable structure on a constantly shifting substrate. The talk will describe Walden's Paths and its Path maintenance subsystem. Further information about the project is located at http://www.csdl.tamu.edu/walden.

Biography

Richard Furuta is a faculty member at Texas A&M University where he is a Professor in the Department of Computer Science, Director of the Hypermedia Research Laboratory, and Associate Director of the Center for the Study of Digital Libraries.

Dr. Furuta's current areas of research include digital libraries, hypermedia systems and models, structured documents, and document engineering. He also has studied applications in computer supported cooperative work, software engineering, visual programming, document structure recognition from bitmapped sources, and management systems for three-dimensional-gesture-based user interfaces. In the area of Digital Libraries, he was one of the founders of the 1994 and 1995 Digital Libraries Conferences, which subsequently became the ACM Digital Libraries series, and later merged with the IEEE-CS series to form the ACM/IEEE-CS Joint Conference on Digital Libraries (JCDL). He was program chair for ACM Digital Libraries 2000, currently serves as Chair of the Steering Committee for ACM/IEEE-CS JCDL, and as an Editor-in-Chief of the Journal of Digital Libraries. In other technical activities, he has been co-program chair for ACM Document Engineering 2002, co-program chair for ACM Hypertext '93, program chair for Electronic Publishing '90, co-program chair of the 1991 DC ACM Chapter annual symposium, Chair of ACM SIGLINK from 1993-1995, member of the ACM SIG Board/SIG Governing Board Executive Committee from 1997-2001, and has served on many other program committees, conference committees, steering committees, and editorial boards.

NSF's latest "Major Research Equipment" project intends to revolutionize how earthquake engineering research is conducted. What will make this possible is an ambitious infrastructure revolving around advanced information technology: very highspeed networking, collaboration technologies, distributed data acquisition, large-scale data storage and management, high-performance computing, and augmented reality. Its success will depend not just on access to these special resources, but on how usable we can make them for engineering researchers. This presentation describes how advances in computer science are being harnessed to change the nature of engineering research. It focuses on the usability challenges that must be addressed so that a national community of researchers can share state-of-the-art laboratory facilities, data, computational models, and research experiences. Examples are shown of what OSU computer scientists are contributing to this national effort.

Biography

Cherri M. Pancake is Professor of Computer Science and Intel Faculty Fellow at Oregon State University. Her previous career involved extensive ethnographic fieldwork, where she applied cross-cultural survey and interviewing techniques to study social change in Guatemalan Indian communities. After earning a Ph.D. in Computer Engineering from Auburn University, Pancake began applying both ethnographic and engineering techniques to the problem of how software tools and Web-based interfaces can more closely match users' intuitions. Pancake is director of the Northwest Alliance for Computational Science and chair of the Parallel Tools Consortium, both collaborative efforts involving computer scientists and scientists from a wide variety of disciplines. She serves as Strategic Advisor on Usability for the San Diego Supercomputer Center, as well as leading the information technology portion of the consortium developing NSF's George E. Brown Network for Earthquake Engineering Simulation. Pancake is a Fellow of both the ACM and the IEEE. She is also currently serving as department head of Computer Science. See www.cs.orst.edu/~pancake and www.nacse.org.

Computational and data Grids couple geographically distributed resources such as high performance computers, workstations, clusters, and scientific instruments. Accordingly, they have been proposed as the next generation computing platform for solving large-scale problems in science, engineering, and commerce.

To date there are very few examples of programming environments that allow legacy applications to be “Grid Enabled”, and thus all Grid demonstrators have been constructed from scratch. Middleware software layers like Globus and Legion are powerful, but they tend to provide a set of low-level primitives which must be called from within the application. This means that at present, in order to build a general Grid application, it is necessary to modify legacy code.

In this seminar I will present two projects in which the goal has been to Grid enable legacy software. The Nimrod project http://www.csse.monash.edu.au/~davida/nimrod.html) has targeted parameter sweep studies in which many independent tasks are distributed to grid resources. Nimrod uses a novel computational economy to enforce a deadline based quality of service system. The GirddLeS project (http://www.csse.monash.edu.au/~davida/griddles.html) targets more general grid applications built from a set of co-operating legacy components. These are linked by a communication protocol called GridFiles.

The architectural design of a software system can support analysis of high-level properties and provide benefits for many software evolution tasks. However, existing tools decouple the architecture of a program from its implementation, allowing inconsistencies to accumulate as the system evolves. Because of the potential for inconsistency, engineers evolving a program cannot trust the architecture to accurately describe the properties or structure of the implementation. This talk presents a new approach: integrating architectural descriptions into an implementation language, and using a type system to ensure that the architecture is consistent with the code. The approach is embodied in the ArchJava language, which extends Java with features that declare the software architecture and data sharing within a system. ArchJava's type system enforces architectural conformance, the property that implementation components communicate only in the ways specified by the architecture. ArchJava is flexible enough to describe hierarchical architectures that may change at run time, and it supports many of the same coding styles and idioms that programmers use in Java. Several case studies applying ArchJava to existing programs provide preliminary evidence that ArchJava is practical and can aid software evolution tasks.

The new OSU Online Catalog & Schedule of Classes was developed using Microsoft ASP.NET and ADO.NET technologies. We will take a look at how these technologies, along with the Object Oriented nature of the .NET Framework, accelerated the development, and facilitated prototyping used to stimulate user feedback during the development process. Some of the key aspects of the architecture of the application will also be discussed. The Online Catalog & Schedule of Classes can be found at http://catalog.oregonstate.edu.

Biography

Mark Clements graduated from Oregon State University with a BA in Spanish in 2001. During his years as a student he worked as a computer lab manager for the College of Business and worked his way into web development there. He is now a Senior Systems Development Engineer for the Business Solutions Group in the College of Business, of which he is a founding member. Mark is a self-taught developer and has been developing .NET applications and services for 2 years.

Bayesian belief networks are being implemented with increasing frequency in ecological modeling. In particular, ecologists are interested in understanding the mechanisms by which stressors to the environment affect the health of ecosystems. Current methodology for fitting Bayes networks involves an assumption of spatial independence that is improbable in ecological settings. In this talk, I will describe some relatively simple Bayesian networks for modeling an ecological system, and describe some of the statistical issues for fitting those networks.

Biography

Dr. Gitelman has been at OSU for 4 years and has a doctorate in Statistics from Carnegie Mellon University. In her thesis research she extended a class of causal models to the setting of hierarchical models. Currently, Dr. Gitelman is working in Bayes networks, which are another class of causal model, but not unrelated to hierarchical models.

Tom Dietterich and Adam Ashenfelter
OSU Computer Science

Recently, Lafferty, McCallum and Pereira introduced the Conditional Random Field as a new model for solving sequential supervised learning problems. Many applications of machine learning can be formalized as Sequential Supervised Learning (SSL). Each training example in SSL has the form (X,Y), where X is a sequence (x1, ..., xT) of items (each typically described by a vector in R^n), and Y is a sequence (y1, ..., yT) of class labels from {1, ..., K}. Typical applications include part-of-speech tagging, information extraction from web pages, and text-to-speech mapping. There are no robust, off-the-shelf methods for solving SSL problems in any commercial or academic statistical or data mining software systems. This talk will provide an introduction to the problem, discuss why the CRF is a good candidate for an off-the-shelf method, and describe our work on applying Friedman's Gradient Tree Boosting algorithm to efficiently and flexibly fit CRF models to the large NETtalk text-to-speech data set. Preliminary performance results will be presented.

Biography

http://www.cs.orst.edu/~tgd

Dr. Dietterich (AB Oberlin College 1977; MS University of Illinois 1979; PhD Stanford University 1984) joined the OSU faculty in January 1985. In 1987, he was named a Presidential Young Investigator for the NSF. In 1990, he published, with Dr. Jude Shavlik, the book entitled Readings in Machine Learning, and he also served as the Technical Program Co-Chair of the National Conference on Artificial Intelligence (AAAI-90). From 1992-1998 he held the position of Executive Editor of the journal Machine Learning. The American Association for Artificial Intelligence named him a Fellow in 1994, and the Association for Computing Machinery did the same in 2003. In 2000, he co-founded a new, free electronic journal: The Journal of Machine Learning Research. He served as Technical Program Chair of the Neural Information Processing Systems (NIPS) conference in 2000 and General Chair in 2001. He currently President of the International Machine Learning Society and he also serves on the Board of Trustees of the NIPS Foundation.

People increasingly use the web not only to browse information, but to perform actions, such as online purchases. Existing web browsers do not provide specific support for recording and reviewing these online actions, however. This support is particularly necessary when something goes wrong and users need to debug the processes they initiate. I will present Woodstein, an agent that monitors and displays user actions on the web. Woodstein explains the details as well as the structure of these processes and also supports debugging by helping users keep track of their own debugging process.

Biography

Earl Wagner is a masters' student at the MIT Media Laboratory. He is interested in software that helps users understand and modify the systems they interact with, especially when something goes wrong. He is currently working with Henry Lieberman to develop Woodstein, an end-user debugger for e-commerce that helps users understand their actions on the web. Before coming the Media Lab, he researched technologies for software development environments and received a B.S. in computer science from the University of California at Berkeley. Homepage web.media.mit.edu/~ewagner

Biography

Rick Lindsley is a software engineer with 20 years experience in Unix and Linux. He's worked on everything from Berkeley 4.2 and 4.3 releases to Solaris to, now, Linux. He spent several years at Tektronix in Beaverton, Oregon, then gained significant experience with multi-processor machines while working for Sequent Computer Systems for nine years. When Sequent was purchased by IBM in 1999, he stayed on board and joined IBM's Linux team. He's currently working in the Linux Technology Center for IBM in Beaverton Oregon. Since then he's given talks at universities, national laboratories, and Linux user groups on topics ranging from what it's like to work for IBM, to issues with SMP locking, to the current state of the Linux kernel. His recent focus is on I/O statistics gathering and process scheduling improvements.

Multithreaded architectures have received considerable attention over the past few years. However, these architectures rely on conventional programming paradigms and require complex runtime transformation of the control-flow programs into dataflow programs, requiring complex hardware to detect data and control hazards, reorder and issue multiple instructions. Our architecture differs from other multithreaded architectures in two ways: i) our programming paradigm is based on dataflow, which eliminates the need for complex runtime scheduling, thus reducing the hardware complexity significantly, and ii) complete decoupling of all memory accesses from execution pipeline. The underlying dataflow and non-blocking models of execution permit a clean separation of memory accesses (which is very difficult to coordinate in other programming models). Data is pre-loaded into an enabled thread's register context prior to its scheduling on the execution pipeline. After a thread completes execution, the results are post-stored from its registers into memory. The instruction set implements dataflow computational model, while the execution engine relies on control-flow like sequencing of instructions. Unlike Superscalars, our architecture performs no (dynamic) Out-of-Order execution and thus eliminates the need for complex instruction scheduling hardware. In this talk I will present the performance comparisons of our architecture with superscalar and VLIW systems for various benchmarks. The results show that our architecture scales better than other systems when more resources are added.

Biography

Krishna Kavi is currently a Professor and the Chair of Computer Science and Engineering department at the University of North Texas, in Denton, Texas. Prior to joining UNT, he held faculty positions at the University of Alabama in Huntsville and the University of Texas at Arlington. He also served as a Program Manager at the National Science Foundation. He was an IEEE Computer Society (CS) Distinguished Visitor (1989-91), editor of the IEEE Transactions on Computers (1993-1997), and editor of the Computer Society Press (1987-1991). His primary research interest lies in Computer Systems Architecture, including dataflow and multithreaded systems, Memory Management, Operating Systems, and Compiler Optimization. His other research interests include Formal specification of concurrent processing systems, Performance Modeling and Evaluation, load balancing and scheduling of parallel programs. He published over 125 technical papers on these topics. He received his B.E. (Electrical) from the Indian Institute of Science, MS and Ph.D. (Computer Science) from the Southern Methodist University.

With the increasing availability of genomics and proteomics tools that identify new genes, proteins, and biological pathways in which they take part, the ability for biologists to move into new areas of the scientific literature and other resources - a.k.a., the bibliome - is essential. Although the bibliome is widely available through the Internet, considerable work is still required to find and synthesize available information. The Text Retrieval Conference (TREC) is an annual forum for researchers in information retrieval (IR) to evaluate their systems and approaches with a common data set. Most work in TREC has focused on the newswire domain, but a convergence of researchers from the biological and IR fields who are interested in enhancing access to information in the genomics domain has emerged. Dr. Hersh, a noted authority in IR in the biomedical domain will describe the new TREC Genomics Track which aims to advance IR systems for genomics researchers and in turn allow them to improve their research.

Biography

Rick Lindsley is a software engineer with 20 years experience in Unix and Linux. He's worked on everything from Berkeley 4.2 and 4.3 releases to Solaris to, now, Linux. He spent several years at Tektronix in Beaverton, Oregon, then gained significant experience with multi-processor machines while working for Sequent Computer Systems for nine years. When Sequent was purchased by IBM in 1999, he stayed on board and joined IBM's Linux team. He's currently working in the Linux Technology Center for IBM in Beaverton Oregon. Since then he's given talks at universities, national laboratories, and Linux user groups on topics ranging from what it's like to work for IBM, to issues with SMP locking, to the current state of the Linux kernel. His recent focus is on I/O statistics gathering and process scheduling improvements.

The advent of multi-agent systems (MAS) has brought us opportunities for the development of complex software that will serve as the infrastructure for advanced distributed applications. During the past decade, there have been many agent architectures proposed for implementing agent-based systems, and also a few efforts to formally specify agent behaviors. However, research on narrowing the gap between agent formal models and agent implementation is rare. In this talk, we present a model-based approach to designing and implementing multi-agent software systems. Instead of using formal methods for the purpose of specifying agent behavior, we bring formal methods into the design phase of the agent development life cycle. During the presentation, we first introduce the formalism called agent-oriented G-net model, which is based on the G-net formalism (a type of Petri nets), to serve as the high-level design for intelligent agents. To illustrate our formal modeling technique for multi-agent systems, an example of an agent family in electronic commerce is provided. Then we show how an existing Petri net tool can be used to detect design errors, and how model checking techniques can support the verification of some key behavioral properties of our agent models. Finally, based on the high-level design, we derive the agent architecture and the detailed design for agent implementation. To demonstrate the feasibility of our approach, we developed the toolkit called ADK (Agent Development Kit) that supports rapid development of application-specific agents for multi-agent software systems.

Related URL: http://www2.uic.edu/~hxu2/PhD/Thesis.html

Biography

Haiping Xu received the BS degree (1989) and the MS degree (1992) in electrical engineering from Zhejiang University, Hangzhou, China, the MS degree (1998) in computer science from Wright State University, Dayton OH, and the Ph.D. degree (2003) in computer science from the University of Illinois at Chicago, IL. From 1992 to 1996, he successively worked with the Ministry of Electronics Industry, Shen-Yan systems Technology, Inc. and Hewlett-Packard Co., as a software engineer, in Beijing, China. He was a research scholar at Nanyang Technological University, Singapore, during the summer of 1996. His research interests include distributed software engineering, multi-agent systems and model-based software development. He is a member of the ACM, IEEE and IEEE Computer Society.

URL: http://www.cs.uic.edu/~hxu1

I explore problems with many "maybes". In experiments with randomized world generation from abductive inference engines, I saw that (often) a few randomly selected worlds of belief yielded as much useful information as searching many more worlds.

This result seemed crazy- a few quick peeks are as good as many hard stares? Yet after much experimentation, I can report that the effect is repeatable in many domains.

After many years exploring this effect this talk can report WHEN we can expect the effect to repeat, WHY the effect happens, and HOW we can use it to dramatically simplify software engineering.

Biography

Dr. Tim Menzies has been working on advanced software engineering techniques since 1986. He is the author of over 150 research papers and currently is the software engineering research chair at NASA's Independent Verification and Validation (IV&V) Facility. Dr. Menzies received his PhD from the University of New South Wales, Sydney, Australia and is a member of the ACM and IEEE.

For more information, see

http://menzies.us
http://menzies.us/me.html
http://menzies.us/papers.html
http://menzies.us/Draftpub.html

Businesses that rely on web services are vulnerable to the problems of those web services. Service contracts and warranties can provide some assurances. However, they provide traditional recourse, rather than timely alerts of impending problems. While electronic commerce has increased the speed of on-line transactions, the technology of monitoring on-line transactions has lagged behind.

To address the problem of web service monitoring, we integrated methods of requirements analysis and software execution monitoring. The resulting system assists analysts in the development of web service requirements monitors.

The work presented here builds on prior research by: (1) building on a goal-based method for obstacle discovery, (2) illustrating the derivation of assigned monitors from obstacles, and (3) automatically deriving web service monitors from high-level requirements descriptions. The framework, and tool, provides an important contribution by demonstrating how distributed concurrent web service transactions can be monitored at the requirements level.

Biography

Dr. William Robinson is an associate professor at Oregon State University (COB/MIS). He has written over 40 academic articles, mostly in the areas of Requirements Engineering and agent support of Electronic Commerce. His journal articles include Communications of the ACM, IEEE Transactions on Software Engineering, Journal of Global Information Management, Concurrent Engineering: Research & Applications, and ACM Computing Surveys. Dr Robinson is secretary of IFIP Working Group 2.9 (Software Requirements Engineering). He is on the steering committee of the IEEE Requirements Engineering Conference series, and was past program chair of the Fourth IEEE Requirements Engineering Symposium (RE'99).

As CPU speeds and counts have grown in recent years, the cost of maintaining symmetric access across the system has greatly increased. To alleviate these costs machines with Non-Uniform Memory Access (NUMA) have appeared.

The 2.6 Linux kernel includes many enhancements in support of NUMA machines. Data structures and macros are provided within the kernel for determining the layout of the memory and processors on the system. These enable the memory subsystem to make decisions on the optimal placement of memory for processes, and the scheduler to factor in node locality when making process scheduling and load balancing decisions. In addition to items that have been incorported into the mainline Linux kernel, there are NUMA features that have been developed that continue to be supported as patchsets.

This talk will cover 2 specific features that have grown out of these development efforts: kernel text replication and user page replication. Both of these features aim to make local copies of read-only data to enhance NUMA performance, but the solutions are radically different.

Biography

Dave Hansen graduated from Purdue University with a B.S. in Computer Science in 2001. He joined IBM's Linux Technology Center in Beaverton, Oregon where he works with the Linux community to increase kernel scalability. His work has run the gamut from SMP locking, to filesystems, to networking, but his current work is on the Linux virutal memory subsystem.

First and foremost, we all [should] know that the real world is continuous and ANALOG (unless you dig deep down to the discrete nature of photons). Despite the fast growth of digital systems today, an analog circuit such as analog-to-digital (A/D) converter is a necessary component in just about all digital systems, as the real-world interface between analog and digital has to be addressed at some point. In the area of analog chip design, the shrinking transistor dimensions (which is good for digital like the pentium processor) is anticipated to create some big problems. This is because small transistors can tolerate only a small amount of voltage stress. A 0.18 micron CMOS process, for example, can only tolerate 1.8 volts. My seminar will briefly review this low-voltage problem; summarize some of the well-known solutions currently in use (and problems associated with these solutions); and new circuit techniques that we have recently developed here at Oregon State University. Some chip implementation results of pipelined A/D converter will be presented. I will also devote about 5-10 minutes of my talk to go over my background and some other research topics at OSU that my research group is currently involved in.

Biography

Prof. Moon received B.S. from University of Washington, M.Eng. from Cornell University, and Ph.D. from University of Illinois, Urbana-Champaign. From February 1994 to January 1998, he was with Bell Laboratories. Since January 1998, he has been with Oregon State University. His research area is in analog and mixed analog-digital integrated circuits.

Energy Systems encompasses the areas of Power Electronics, Machines and Drives in Industrial Processes, and Power Systems including Alternative Energy Interfaces. This seminar will present an overview of the Energy Systems research at OSU, which has the central theme of "Improved Power Processing", with a wide variety of interesting projects and applications. Specifically, my research expertise has focused on Power Electronics (power supply design, converters and multi-level inverters), Power Quality and Renewables (including utility interface issues and filtering/conditioning approaches) and Adjustable Speed Drives (application issues including overvoltage, bearing currents and electromagnetic interference, as well as ride-through). This seminar will also describe the Motor Systems Resource Facility (MSRF) - the highest power Energy Systems laboratory in any university in North America (including a 750kVA dedicated power supply, comprehensive testbeds up to 300hp, and a 120kVA fully programmable source). www.ece.orst.edu/~avj, eecs.orst.edu/msrf/

Biography

Dr. von Jouanne is in her 9th year as a professor in the School of EECS at OSU. She is in the Energy Systems group working primarily on power electronic converters and industrial drives, power quality and renewables. She is the Co-Director of the Motor Systems Resource Facility, a registered professional engineer, and was the recipient of the 2000 IAS Outstanding Young Member Award. Dr. von Jouanne received her Ph.D. degree in Electrical Engineering from Texas A&M University where she also worked with Toshiba International Industrial Division. She received her M.S. and B.S. degrees in Electrical Engineering with a Minor in Mathematics from Southern Illinois University. She has been interviewed for the internationally-syndicated NPR program "51%", highlighting women who have impacted society, especially through science and technology. She has also been selected by the National Academy of Engineering to be profiled as one of their "Celebrated Women Engineers". www.ece.orst.edu/~avj

A brief overview of ongoing Materials & Device research at OSU is presented. This work is an interdisciplinary effort involving researchers in the Departments of Chemistry, Chemical Engineering, EECS, and Physics. Research topics to be surveyed include transparent electronics, low-cost electronics, photovoltaics, and intelligent luminescence.

Biography

J. F. Wager is a Professor in the School of EECS at OSU whose research specialization is in the area of solid-state materials and devices.

http://eecs.oregonstate.edu/research/members/wager/index.html

Reinforcement Learning (RL) is the study of systems that improve their performance at some task by taking actions and receiving rewards and punishments. The rewards are propagated backwards in time so that the system prefers actions that lead to large rewards in the long run. Reinforcement Learning is an active area of research with many theoretical and practical results. In this talk, I focus on scaling RL to domains with large state spaces through abstraction hierarchies and function approximation. I'll also outline some ideas on scaling RL to large action spaces and to relational domains.

Biography

Prasad Tadepalli has an M.Tech from Indian Institute of Technology, Madras, and a Ph.D from Rutgers University, both in Computer Science. He has been teaching at Oregon State University since 1989. His main area of research is Machine Learning, including Reinforcement Learning, Relational Learning, and Computational Learning Theory with applications to classification, real-time scheduling, and information extraction. http://web.engr.oregonstate.edu/~tadepall

The computing landscape is undergoing a dramatic transformation as computing devices become more ubiquitous, network-enabled and ready to communicate. This trend is not only impacting the way we use computers, but also the way we build software for them. This talk centers on adaptive approaches to distributing content and to performing distributed computations. We first motivate adaptive computing by presenting “Motion,” a mass video distribution network that uses adaptive techniques to deliver video daily to two million users out of “recycled” bandwidth. This approach improves the user experience and significantly reduces the cost of delivery. We’ll then focus on tuple-space models as a basis for adaptive computation; our early work with the Linda coordination language at Yale University suggests that these models have the right characteristics for adaptive computation, while later work with Sun Microsystems on JavaSpaces suggests their utility in a spontaneous network environment. We’ll also introduce and discuss the role of distributed data structures in these computations and their growing presence in today’s Internet. Finally, we will briefly touch on Lifestreams, an interface for the network environment (the speaker’s thesis work), and then suggest future directions in adaptive computing based on current Internet infrastructure and trends.

Biography

Dr. Eric Freeman is currently Director of Engineering for the Walt Disney Internet Group, where he directs broadband and wireless strategy, technology, and product development. Prior to joining Disney, Eric co-founded Mirror Worlds Technologies, Inc., a company that produces a commercialized version of his Ph.D. work. Eric was also previously on faculty at Yale University as a Research Affiliate, where he worked closely with Sun Microsystems and focused on distributed computing using Java and Jini technologies. As a result of this work, Eric co-authored JavaSpaces Principles, Patterns and Practice, the official Sun book on JavaSpaces. Eric received a Ph.D. from Yale University in 1997 for his work on the Lifestreams system, one of the first research systems to suggest an alternative for the desktop metaphor. He was recognized for this work in 1999 by MIT\'s Technology Review as one of the top 100 young innovators.

Eric’s research interests include distributed computing, Internet technologies, information systems and programming languages.

The TaskTracer Project is a newly initiated computing research project at Oregon State University that is investigating the possibilities of a desktop software system that will track in detail how knowledge workers complete tasks, and intelligently leverage that information to increase efficiency and productivity. The goal is to create a system with four capabilities: more efficient task-interruption recovery, individual knowledge management, workgroup knowledge management, and within-workgroup workflow detection and analysis. The proposed system will operate in the Microsoft Windows environment, tracking most interactions with desktop applications as well as tracking phone calls. A central challenge involves “unweaving” the threads of events that are affiliated with different tasks that are being performed concurrently. Our approach will combine creative user interfaces and machine learning to perform this unweaving. Other challenges include identifying “significant” events, summarizing tasks or subtasks, predicting likely steps and resources, and detecting cross-user workflow. The TaskTracer project at Oregon State is part of the Management of Knowledge-Intensive Dynamic Systems (MKIDS) initiative funded jointly by the National Science Foundation and the Intelligence Community.

Biography

Jon Herlocker (BS Lewis & Clark College 1994; PhD Minnesota 2000) is Assistant Professor of Computer Science at Oregon State University. Dr. Herlocker's research work centers on integrating intelligence and usability into information systems. At the University of Minnesota, he was part of a group of researchers that developed a collaborative filtering-based recommender system which was licensed to Net Perceptions Inc, and was the lead designer and developer for the MovieLens web-based movie recommender (www.movielens.org). He has published a number of scientific papers on algorithms and systems for collaborative filtering, and has been awarded the prestigious NSF Faculty Early Career Development (CAREER) award for his work in that area. Dr. Herlocker is a member of the ACM and ACM SIGCHI, the IEEE and the Computer Society, and is the faculty advisor the local student chapter of the ACM. http://cs.oregonstate.edu/~herlock/

Through a series of examples, I will demonstrate presentation techniques that enhance or detract from the message you are trying to deliver. I will discuss how much information to put on a single slide, how slides should relate to one another, subliminal messages slides may convey, when to be silent, the use of gestures, question-answering strategies, and more.

Biography

Professor Michael J. Quinn earned an honors B.S. in mathematics from Gonzaga University in 1977, an M.S. in computer sciences from the University of Wisconsin-Madison in 1979, and a Ph.D. in computer science from Washington State University in 1983. From 1979 to 1981 he worked for Tektronix, Inc. as a software engineer. He was an assistant professor of computer science at the University of New Hampshire from 1983 to 1989, before joining Oregon State University in 1989. He served as interim Head of Computer Science in 1997-1998 and Head of Computer Science from 1998-2002.

Dr. Quinn is the author of dozens of refereed publications in the area of parallel computing. He has also authored or co-authored three books on parallel computing. His fourth book, Parallel Programming in C with MPI and OpenMP, will be published by McGraw-Hill in June 2003. Dr. Quinn is a member of the ACM, the IEEE, the IEEE Computer Society, and Computer Professionals for Social Responsibility. He has received the "Golden Core" service award from the IEEE Computer Society.

The business group SMS (for Secure Mobile Solutions) of Infineon Technologies AG is the fusion of the former business groups WS (for Wireless Solutions) and CC (for Security and ChipCard ICs). I will shortly glimpse over their activities and their products concentrating on those inherently related with security. After a short defense on the importance of merging mobility and security in todays days, I will present the engineering difficulties when developing real-world security products. In addition to that, I will also show how to overcome those security challenges with novel ideas and techniques. All in all, I will guide the audience through the development of a modern high-security micro-controller as used in high-end ChipCard ICs or government crypto-chips.

Biography

Jean-Pierre Seifert received the diploma degree in Mathematics and Computer Science from the Johann-Wolfgang Goethe University at Frankfurt on the Main in 1995. In 1999 he received the Doctorate degree in Mathematics also from the Johann-Wolfgang Goethe University. During 1995 and 2000 he had several visiting positions at the MIT, ETH Zuerich and the Queensland University of Technology. In 2000, he joined Infineon Technologies AG. In 2002 he received Infineon's Inventors of the year" Award for his outstanding numbers of filed patents. Currently he is a Principal Security Research Scientist within the business group Secure Mobile Solutions (SMS) of Infineon Technologies. There, he is leading the "Security Concepts and Innovations" group for the whole SMS business group.

In 2003 he received an offer as a Full Tenure Professor chairing the research field Cryptographic Engineering within the Computer Science Department at the Universitaet Saraviensis (Saarbruecken, Germany). In addition to his large number of patents, he has published several papers on Algorithmic Number Theory, Approximation of NP-hard Problems, Complexity Theory, Computer Arithmetic, Computer Security, Cryptography, Digital Circuit Design, Hardware Design, Lattices in Cryptography and Cryptanalysis, Side channel Analysis, Tamperproof hardware & software design, and also on Quantum Computing.

Many will agree that the fork is perhaps the ultimate recourse when developing extensions to open source and free software projects. Given turnaround time of integration with many large projects, many developers resort to maintaining what are effectively their own forks.

The GUFT Unified Forking Toolkit aims to ease the burden of maintaining common "splinter forks" and aims to make "separate forks" more prevalent. The project's ultimate goal will be to someday see a "Fork Our Distribution" link available from every open source operating system vendor website.

The work will be accomplished with the following set of integrated toolkits:

* GOLT Open-Lean Toolkit (Software Engineering)
* GURT Unified Repository Toolkit (Version Control)
* ARMPIT Recursive Meta-Package Integration Toolkit (Package Management)
* BUZCUT Utterly Zero Click Upgrade Toolkit (Site Management)

Together, these toolkits will provide a comprehensive solution for creating, maintaining, and redistributing splinter or separate forks of complex software projects. In the end, these should allow a single individual or small team to manage several moderately unique distributions. With such a toolkit in hand, companies will be able to leverage forking in tandem with a diversified branding strategy, resulting in what can only be described as "software development franchises".

This talk will explain the rational for creating such a toolkit along with the requirements used to develop the GUFT prototype and its underlying architecture.

Biography

Zach Welch formed Superlucidity Services, LLC to provide consulting services for open source software. After porting Gentoo Linux to ARM, he formed The Zynot Foundation (www.zynot.org) which helped him develop ideas for beneficial forking. He is currently developing a prototype of GUFT and evangelizing the OSU Open Source Lab (www.osuosl.org).

Humanoid robots and agents are becoming increasingly prevalent in a wide variety of applications, ranging from animated human characters to physically embodied humanoid robots. Such humanoids will eventually act autonomously, interacting with humans through collaboration rather than explicit programming. Manually designing and implementing control mechanisms for autonomous humanoids can be a complicated and time-consuming process that lacks scalability and is not suitable for non-technical developers. The aim of my research is to address this problem by automatically designing and implementing modular humanoid skills derived from abilities demonstrated by human beings. In this talk, I describe data-driven methods for unsupervised learning of such skills from human motion.

These methods address two main problems: 1) how can human motion data be analyzed to produce modular humanoid skills and 2) how can suitable human motion data for data-driven analysis be captured. Each of these projects utilizes or extends multidimensional scaling techniques to uncover structure within data produced by human motion in a model-free fashion.

The primary theme of my research is to derive humanoid skills from motion data of humans behaving ``naturally''. An underlying philosophy of this work is that motion capture should require little or no instrumentation such that a subject can move in an unrestricted manner while performing tasks. Towards this end, this talk will describe a method for Kinematic Model and Motion Capture (KMMC). KMMC automatically estimates the kinematic model and joint angle motion from video sequences of multiple calibrated cameras. Furthermore, KMMC can be applied to subjects with arbitrary genus 0 (tree-structured) kinematics. The KMMC method is complimented by Performance-Derived Behavior Vocabularies (PDBV) for deriving ``behavior vocabularies'' from unlabelled motion data of a human performing multiple activities. PDBV uses a spatio-temporal method for multidimensional scaling to uncover behaviors that underlie motion data. Uncovered behaviors can then be realized as nonlinear dynamical systems in the joint angle space of the humanoid. Such dynamical systems are used as predictors for a variety of applications. This talk will demonstrate the utility of behavior vocabularies for humanoid motion synthesis, classification, and imitation.

Biography

Chad Jenkins recently completed his Ph.D. in the Computer Science Department and the Center for Robotics and Embedded Systems at the University of Southern California, under the supervision of Prof. Maja Mataric'. He earned his B.S. in Computer Science and Mathematics at Alma College (1996) and M.S. in Computer Science at Georgia Tech (1998). His research interests include humanoid robotics, machine learning, computer animation, computer vision, and autonomous agents. Recently, Chad presented his research at the International Robots and Systems Conference (IROS), Autonomous Agents and Multi-Agent Systems (AAMAS), and Computer Vision and Pattern Recognition (CVPR).

Low-jitter clock generation and clock/data recovery (CDR) blocks are important components required for reliable operation of high-speed synchronous systems such as microprocessors and data communication links. These blocks must contend with several noise sources that can adversely impact performance. This talk presents a mixed PLL/DLL architecture for low-jitter clock generation and to be used as a core component for clock/data recovery. This architecture merges the characteristics of PLLs and DLLs via an interpolator in order to easily adjust loop dynamics in response to different noise environments.

Biography

Gu-Yeon Wei joined Harvard University in January 2002 as an Assistant Professor of Electrical Engineering in the Division of Engineering and Applied Sciences. Prior to joing Harvard, he spent 18 months at Accelerant Networks in Beaverton, Oregon as a Senior Design Engineer. Professor Wei received his BS, MS, and Ph.D. degrees all from Stanford University in 1994, 1997, and 2001. His current research interests are in the area of mixed-signal VLSI circuits and systems design for high-speed/low-power wireline data communication, low-jitter clock generation, energy-efficient computing devices for sensor networks, and biosensor applications.

In this talk I will give an overview of empirical software engineering. Then I will discuss two specific research projects I have been working on. The first project is the development of a qualitative methodology for conducting empirical software engineering research. The second project deals with software engineering for high performance computers.

Empirical software engineering is useful or building support for hypotheses. But, the context changes from one study to the next, so it is not always clear how to compare their results. Various results from isolate studies are not as useful as the deeper conclusions that can be drawn when those results are analyzed together. The methodology allows researchers to identify variables, generate hypotheses and build support for those hypotheses. The methodology consists of the following steps, which will be discussed in more detail:

1) Gather expert opinion about potential variables;
2) Develop and refine hypotheses for those variables based on existing data;
3) Run new studies to verify the hypotheses.

After giving an overview of the methodology, I will describe its use for identifying and building support for hypotheses about the relationship of an inspector’s background and experiences to his or her effectiveness during an inspection. The goal of this investigation was to explain the wide variation in the performance of different inspectors.

Finally, I will give an overview of the work I have been doing concerning software engineering for high performance computers. Currently there is very little quantitative understanding of the effort required to develop software for high performance computers or of the tradeoffs between development time and execution time. I will describe a series of planned studies and how we plan to use the results of those studies to generate hypotheses and better understand these issues.

Over the years, probability theory, utility theory and decision theory have emerged as principled frameworks for the design of systems that can adapt, learn and robustly strategize in the presence of uncertainty. In particular, partially observable Markov decision processes (POMDPs) combine these theories into one general framework that can naturally model a wide range of real-word sequential decision problems.

Unfortunately, the considerable expressivity of POMDPs generally results in solution algorithms that are computationally infeasible. In this talk, I will first introduce POMDPs and discuss some important sources of intractability, including the curse of dimensionality as well as the representation of policies and value functions. I will then present a new belief compression technique that mitigates the curse of dimensionality and a bounded policy iteration algorithm that alleviates the complexity of policy search. Both of these approaches allow the solution of large POMDPs more quickly than existing algorithms. Finally, I will discuss the combination of these methods, allowing approximate solutions for problems with belief spaces of dimensionality several orders of magnitude larger than previously addressed in the literature. I will illustrate the application of POMDPs and the above algorithms with a system designed to help elderly persons with various forms of dementia (e.g., Alzheimer's disease) to carry out simple daily tasks.

The current ``best-effort'' Internet, however, does not guarantee Quality of Service (QoS) such as minimum bandwidth, packet loss rate, and delay which are critical to many multimedia applications. As such, many significant challenges remain to design and deploy delay sensitive multimedia applications over the Internet effectively. In this talk, I will present the path diversity (PD) framework for concurrent media streaming to a receiver using multiple routes. Without requiring QoS, the PD framework improves the quality of the streamed media via multiple routes created using either multiple senders or relay nodes, in order to increase available bandwidth, reduce packet loss and delay. The PD framework combats packet loss, delay, and insufficient bandwidth for pre-recorded streaming media by sending packets simultaneously from multiple senders to a single receiver. For interactive and live streaming applications, the PD framework allows a single sender to send packets simultaneously on both default and redundant paths to the receiver. Within the PD framework, I will present a transport protocol to synchronize the simultaneous media streaming to receiver via multiple routes. In particular, the protocol employs the rate allocation and packet partition algorithms. The rate algorithm determines the sending rate on each route in order to minimize the packet loss, while the packet partition algorithm ensures each packet is sent by one and only one sender and at the same time, minimizes the startup delay. I will show theoretically and experimentally that using Forward Error Correction (FEC) in streaming the media simultaneously over multiple mostly independent routes at appropriate sending rates is more effective than using FEC with the traditional uni-path streaming. I will also show the performance gain of coupling Multiple Description Coding (MDC) video with the PD framework over the traditional Single Description Coding (SDC) video using the traditional uni-path approach.

Biography

Thinh Nguyen received his B.S. from University of Washington in 1995, his M.S. and Ph.D. from U.C. Berkeley in 2000 and 2003, respectively. His doctoral work focused on multimedia streaming over the Internet, using the path diversity framework in conjunction with network protocols, and source and channel coding techniques. His research interests include multimedia networking, computer networks, signal processing, machine learning, data analysis and data mining. Prior to graduate school, he worked on computer architecture, advanced graphics and visualization algorithms at Intel Microcomputer Research Lab and at Microsoft Corp. He has won a number of awards including the “Best Paper” award at IEEE Packet Video Workshop for his work on distributed video streaming with forward error correction. More information about him can be found at:
http://www-video.eecs.berkeley.edu/~thinhq

In this talk I'll describe novel learning algorithms for temporal, relational data and their application to trainable video interpretation. I extend an existing visual-event recognition system, Leonard (Siskind 2001), with two new learning components:

1) A new relational sequential inference method that learns mappings from relational observation sequences to relational state sequences. Using this method, we infer force-dynamic world models from raw video, with results comparing favorably to pre-existing hand-coded model reconstructors.

2) A supervised learning method for logical event definitions written in terms of the force-dynamic models constructed in (1) above. I give a specific-to-general learning algorithm that learns definitions that empirically outperform published definitions written by a human expert.

Biography

Alan Fern is a Ph.D candidate in Computer Engineering at Purdue University. He received an M.S degree from Purdue in 2000 and a B.S from the University of Maine in 1997. While at Purdue, Alan received an NSF Graduate Research Fellowship. His primary research interests are in machine learning, data mining, and automated planning/control. He is particularly interested in developing algorithms that leverage rich knowledge representations for learning and discovery in highly structured data.

This talk presents techniques for reducing the power dissipation and chip area of analog filters. The emphasis is on filters used for channel selection in wireless receivers. We first discuss a scheme that uses multiple filtering paths, each optimized for portion of the required total dynamic range. Since the individual filtering paths have small dynamic range they require small power dissipation and chip area. The filtering paths operate all the time allowing the overall system to provide undisturbed output during range-switching, contrary to conventional AGC-filter schemes that give disturbances every time the gain changes. The second part of the talk presents an optimization algorithm for maximizing the dynamic range of an arbitrary filter under a fixed power dissipation constraint. The algorithm uses a simplified Volterra series representation to describe distortion. Closed form expressions are derived for the optimal allocation of power dissipation among the elements of the filter. We next present experimental results from a chip that has been implemented using the proposed techniques. The talk will conclude with a brief overview of the activities of the Communications Circuits Lab of Intel Corporation, Hillsboro, OR. These activities include fully integrated CMOS transceivers, VCOs, synthesizers, very high frequency circuits, etc.

Biography

Yorgos Palaskas received the Diploma in electrical and computer engineering from the National Technical University of Athens, Greece, in 1996, and the M.S. and Ph.D. degrees, both in electrical engineering, from Columbia University, New York, in 1999 and 2002, respectively. His Ph.D. work was in the area of syllabic companding filters for wireless applications. During the summer of 1999 he was with Texas Instruments, NJ, where he worked on disk drive electronics. During the summers of 2000 and 2001 he worked at Agere Systems, New Jersey (formerly Bell Labs), doing research on integrated IF filters. Since January 2003 he has been a Sr. Design Engineer at the Communications and Interconnect Technology Lab of Intel Corporation, Hillsboro, OR, where he is working on wireless transceiver architectures. He holds one US patent and has several others pending.

Attacks against Internet routing are increasing in number and severity. Contributing greatly to these attacks is the absence of origin authentication: there is no way to validate if an entity using an address has the right to do so. This vulnerability is not only a conduit for malicious behavior, but indirectly allows seemingly inconsequential misconfigurations to disrupt large portions of the Internet. This talk discusses the semantics, design, and costs of origin authentication in interdomain routing. A formalization of address usage and delegation is presented and broad classes of cryptographic proof systems appropriate for origin authentication are considered.

The costs of origin authentication are largely determined by the form and stability of the served address space. However, prior to this work, little was known about the relevant characteristics of address use on the Internet. Developed from collected interdomain routing data and presented in this talk, our approximate delegation hierarchy shows that current IP address delegation is dense and relatively static. One notable result shows that as few as 16 entities are the source of 80% of the delegation on the Internet. We further show via simulation that these features can be exploited to efficiently implement Internet-scale origin authentication. The talk is concluded with a presentation of thoughts on major problems in routing security and other related future work.

Biography

Patrick McDaniel is a Senior Technical Staff Member of the Secure Systems Group at AT&T Labs-Research. He received his Ph.D. from the University of Michigan in 2001 where he studied the form, algorithmic limits, and enforcement of security policy. Patrick's recent research efforts have focused on security management in distributed systems, network security, and public policy and technical issues in digital media. Patrick is a past recipient of the NASA Kennedy Space Center fellowship, a frequent contributor to the IETF security standards, and has authored many papers and book chapters in various areas of systems security. Prior to pursuing his Ph.D. in 1996, Patrick was a software architect and program manager in the telecommunications industry.

Ad hoc networks form opportunistically and change rapidly in response to the movement of mobile hosts. These networks present new software engineering challenges, especially given the increasing demand for applications tailored to a wide variety of domains. Applications executing in ad hoc networks need to react continuously and rapidly to changes in operating conditions and must adapt their behavior accordingly. Such applications fall in the category of context-aware computing, a field that has received much recent attention. Much of the current work on context-aware computing is limited to presenting only specific types of context information, e.g., location or time. In addition, current context-aware systems often rely only on information directly available to an application via context sensors on the local host. In this talk, I introduce a novel perspective on context-awareness in which the context includes, in principle, any information available in the ad hoc network but is restricted, in practice, to specific projections of the overall context. I will present the design and implementation of a new middleware model that delivers this notion of context to the application programmer. Specifically, the middleware introduces the ability for particular tasks to operate over dynamically and declaratively specified abstract views of the global context. Despite the fact that interactions among hosts in the underlying network are transient and disconnections are frequent, the middleware continuously maintains the desired contextual information. I also present novel network protocols that provide the context-maintenance required to support the middleware’s implementation. The resulting context-aware middleware eases the software engineering challenges encountered when programming applications for ad hoc mobile environments.

Biography

Christine Julien is a D.Sc. candidate in Computer Science and Engineering at Washington University in Saint Louis. She holds a Graduate Research Fellowship from the National Science Foundation. Christine received her M.S. in Computer Science in 2003 and her B.S. in 2000, both from Washington University. Her research interests include software engineering for mobile computing, formal methods for distributed systems, and network algorithm design.

One of the most exciting things about working in the field of scientific and engineering visualization is that there are no rules. Anything that turns data into information-rich understanding and insight is fair game. Thus, visualization is a big bag of tricks. At UCSD’s Xtreme Visualization Lab, we specialize in inventing and using new tricks. And the more tricks, the more varied and effective the collaborations we can pursue. . This talk will discuss four tricks from the bag, and the types of collaborations that they have allowed us to foster. The four are:

Desktop interactive volume rendering – The XVL’s Volume Explorer program (vx) allows interactive direct volume manipulation on the desktop. It can do this with datasets as large as 1 billion voxels. This has allowed us to visualize many 3D datasets that we would otherwise not have touched. This includes new medical and fluids data, as well as non-traditional volumes such as geophysics and a Confederate submarine.

Color 3D models – The XVL’s Center for Visualization Prototypes manufactures color models when they can help the understanding of complex datasets. I will discuss our most recent projects, which include human body parts to investigate different medical treatments, 3D ultrasounds, planets, and a 3D model of the smoke flow from last Fall’s Southern California fires.

Visualization using graphics hardware shaders – Graphics Hardware Shaders are one of the most exciting things to happen to computer graphics in years. Hardware Shaders allow a developer to write his or her own graphics code and have it executed on the graphics board as part of the display process. This portion of the talk will discuss what this is and what it means to future visualization opportunities.

A True 3D volume server – True 3D displays have reached the point where they can be useful for real applications, particularly group collaboratories. We are in the process of setting one of these up as a long-distance 3D display server to enable groups of collaborators to more effectively work together, especially with volume datasets.

A holy grail of component-based software engineering is "write-once, reuse everywhere". However, in modern distributed, component-based systems supporting emerging application areas such as service-oriented e-business, this is difficult. Non-functional requirements (related to quality-of-service (QoS) issues such as security, reliability, and performance) vary with deployment context, and sometimes even at run-time, complicating the task of re-using components.

I will present a middleware-based approach to managing changing requirements of components. Functional component interfaces are decorated with links to independently evolvable, declarative assertions (policies) that advertise non-functional capabilities. Middleware enhancements match, interpret, and mediate requirements of clients and servers at deployment time and/or runtime. A programming model for extending existing applications with components to implement new application level protocols or features will also be presented. This approach offers the software engineer a flexible solution to handle changing QoS requirements in a well-structured manner.

Biography

Eric Wohlstadter is a Ph.D. candidate at the University of California - Davis. His work focuses on improving software development in the face evolving and dynamically changing requirements. He has designed a new middleware architecture in his thesis work and jointly with IBM Research to manage software composition in service-oriented systems.

Scaling of MOS transistors to 0.10 ƒÝm and below has made the implementation of CMOS circuits operating at 20 GHz and higher feasible. At 24 GHz, a wavelength of electro-magnetic waves in free space is 12.5 mm and in silicon it is 3.7 mm. This means a quarter wave antenna needs to be only ~ 3 and 0.9 mm in free space and silicon. These in conjunction with the increases of chip sizes to ~ 2 cm x 2 cm have made the integration of antennas for wireless communication possible. Integrated antennas could potentially be used to relieve the bottleneck associated with global signal distribution inside integrated circuits including a reduction of clock skew. Integrated antennas could also be used for data communication between integrated circuits to lower the I/O pin count, thus reducing the form factor and packaging costs. When integrated with sensors and power sources, a transceiver with integrated antennas could provide a communication link for sensor network nodes. The nodes can be the size of a grain of rice and sufficiently inexpensive that they may be disposable. Such nodes could help to accelerate the realization of the Smart Dust vision. This seminar will review the status of key technologies required to implement these interconnect systems as well as challenges and potential solutions. This seminar will also discuss the paths for signal propagation, performance of integrated antennas on 10-20 ƒÇ-cm silicon substrates commonly used for CMOS and BiCMOS technologies, circuits which could be implemented in mainstream CMOS technologies for this type of applications, and demonstrations of wireless interconnects. The key challenges including the effects of metal structures associated with integrated circuits, packaging and heat removal, and interaction between transmitted and received signals, and nearby circuits will be discussed.

Biography

Kenneth O received his S.B, S.M, and Ph.D degrees in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology, Cambridge, MA in 1984, 1984, and 1989, respectively. From 1989 to 1994, Dr. O worked at Analog Devices Inc. developing sub-micron CMOS processes for mixed signal applications and high speed bipolar and BiCMOS processes for RF and mixed signal applications. He is currently a professor at the University of Florida, Gainesville. He was also the Chief Technology Officer of the Global Communications Devices Inc. between 2002 and 2003. His research group (Silicon Microwave Integrated Circuits and Systems Research Group) is developing circuits and components required to implement analog and digital systems operating between 1 and 100 GHz using silicon IC technologies. The group is currently composed of 18 graduate students. In 2001, he was a visiting professor in the Department of Electrical Engineering and Computer Science at MIT. He was the general chair of the 2001 IEEE Bipolar/BiCMOS Circuits and Technology Meeting (BCTM). He has served as an associate editor for IEEE Transaction on Electron Devices from 1999 to 2001. Dr. O has also served as the publication chairman of the 1999 International Electron Device Meeting. He has authored and co- authored ~ 100 journal and conference publications, as well as holding five patents. Dr. O has received the 1995, 1997, and 2000 IBM Faculty Development Awards, 1996 NSF Early Career Development Award, and 2004 University of Florida Doctoral Dissertation Advisor/Mentoring Award.

The affordance of large personal screens is different from the affordance of regular sized screens. By extending further into the user's visual periphery, large screens address the peripheral vision of the user and result in a natural distinction of screen space into focus and context regions. In this talk, we present a selection of interaction and visualization techniques that help users make better use of the additional screen resources offered by large screens and deal with some of the challenges brought up by these screens. Focus plus context screens turn regular screens into large screens by adding low-resolution peripheral screen space. High-density cursor is a visualization technique that helps users keep track of the mouse cursor while in the user's periphery; Drag-and-pop is an interaction technique that provides users with access to peripheral screen content. Finally, we will investigate in how far lessons learned from large displays transfer back to regular and small screen devices. We will use the example of Halo, a visualization technique that supports spatial cognition by adding a virtual periphery to small-screen devices.

Biography

http://www.patrickbaudisch.com/biography

We have developed a framework for web-based GIS/database (WebGD) applications. Each of these applications allows its user to insert, update, delete, and query location-based data with the map interface displayed by a web browser. Our framework allows a web-based GIS application to be created and maintained easily.

We have built our WebGD framework by using only open source software packages. PostgreSQL, an object-relational database and PostGIS together manage the geospatial data. PostGIS, which is an extension to PostgreSQL, allows PostgreSQL to be used as a backend spatial database for GIS applications. MapServer developed at University of Minnesota generates maps to be displayed on a web browser by using geospatial data stored in a PostgreSQL database. Web pages, including the one that displays the map, are generated by server-side scripts written in PHP.

We also have a form generator that automatically creates from a database schema a set of forms for database access. The WebGD framework and the form generator in combination allow a Web-based GIS/database application to be developed rapidly.

Some of our applications can be accessed at http://yukon.een.orst.edu.

Biography

Toshimi Minoura obtained his Ph.D. degree by working on concurrency and recovery schemes for distributed database systems. He then designed and implemented active object systems. Currently, he is working on web-based GIS/database applications for environmental impact assessment and conservation.

Vector field design on surfaces is necessary for many graphics applications: examplebased texture synthesis, non-photorealistic rendering, and fluid simulation. In this talk, I will present our vector field design system for surfaces that allows a user to create a large variety of vector fields with relatively little effort. Furthermore, the system allows the user to control the number of singularities in the vector field and their placement.

Our system combines basis vector fields to make an initial vector field that meets the user's specifications. The initial vector field often contains unwanted singularities, which cannot always be eliminated due to the Poincaré-Hopf index theorem. To reduce the visual artifacts caused by these singularities, our system allows a user to move a singularity to a more favorable location or to cancel a pair of singularities. These operations provide topological guarantees for the vector field in that they only affect the user-specified singularities. Other editing operations are also provided so that the user may change the topological and geometric characteristics of the vector field.

At the end, I will show the results of applying our vector field design system to several applications: example-based texture synthesis, painterly rendering of images, and pencil sketch illustrations of smooth surfaces.

Biography

Mr. Eugene Zhang is a Ph.D. candidate at Georgia Institute of Technology, where he has been studying Computer Graphics and Visualization with the emphasis on topologybased analysis on surfaces for image synthesis. His works on "feature-based surface parameterization and texture mapping" and "vector field design on surfaces" are two such examples. Prior to coming to Georgia Tech in 1999, He was a database developer and a group manager at Datastream System Inc. from 1995-1999, customizing Computerized Maintenance Management Software. He received a M.Sc in Mathematics in 1994, and a M.Sc in Computer Science in 1995, both from Ohio State University.

Model checking is a technique for systematically searching the possible behaviors of a system for certain kinds of errors. Model checking has been successfully used to analyze hardware, and it has reached a mature point where companies such as Intel used it in their processor development process. Since then, there have been a number of significant efforts to use model checking for analyzing behavioral properties of software, and they have achieved varying degrees of success. Despite the progressive advances of software model checking in the past few years, two main obstacles still hinder its adoption:

(1) the system modeling problem, and
(2) the state-space explosion problem.

In the context of model-checking Java programs, Bogor addresses obstacle (1) by providing support in its modeling language for high-level features commonly found in Java. Bogor addresses obstacle (2) by including reduction techniques specially designed for Java programs.

The Bogor modeling language extension facility is used to incorporate domain-specific abstractions such as an abstraction of the CORBA real-time event channel to address obstacle (1) when model checking designs of Boeing's avionics systems. To address obstacle (2), we developed a divide-and-conquer, domain-specific state-space exploration that leverages state-space properties of quasi-cyclic systems (e.g., real-time avionic systems), thus, it significantly reduces the memory requirement for model checking such systems. I will then conclude by summarizing my research contributions and their impacts.

Bogor is being used as the core model-checking engine in several research projects at Kansas State University (KSU): (1) the Bandera project, a project to provide tool supports for model checking Java programs, (2) the Cadena project, a project to provide tool supports for designing, specifying, and analyzing event-driven component-based distributed systems such as avionics systems, and (3) the SpEx project, a project to provide a variety of useful specifications suitable for analysis of dynamic software using model checking, testing, etc. In addition, Bogor is currently being used in a number of research projects outside of KSU including University of Massachussets, University of Nebraska-Lincoln, Queens University, Williams College, Vanderbilt University, University of Santa Cruz, and EPFL. Finally, we believe that Bogor is an excellent pedagogical vechicle for exploring classic and state-of-the-art model-checking technology, and we are developing a variety of instructional materials including course notes, lecture slides, course projects, and other supplemental online materials to support the use of Bogor in both undergraduate and graduate education.

Biography

Robby is a Ph.D. Candidate at Kansas State University under the supervision of Dr. John Hatcliff and Dr. Matthew B. Dwyer. His research interests lie at the intersection of software engineering and programming languages. He is the designer and the developer of the Bogor software model checking framework. He is involved in the Bandera project, the Cadena project, and the SpEx project. In the past, he worked on specification languages, abstraction techniques, and counter-example visualization for software model checking.

In this talk, we present two spectral methods for the analysis of naturally-occurring datasets: (1) Sparse Principal Component Analysis (S-PCA) to extract features that arise at multiple-scales; and (2) EigenCuts to cluster items in a dataset using pairwise similarities between the elements. In this work, we focus on datasets consisting of images of faces, eyes and hand gestures.

S-PCA is based on the discovery that natural images exhibit structure in a low-dimensional subspace in a local, scale-dependent form. It is motivated by the observation that the traditional PCA approach does not typically recover such representations, due to its single minded pursuit of variance. In fact, it is widely believed that the analysis of second-order statistics alone is insufficient for extracting multi-scale structure from data. Instead of using higher-order statistics, we show that S-PCA can resolve second-order statistics with suitably constrained basis directions and generate data-specific multi-scale representations. Compared to other methods, S-PCA provides an intuitive model and hence a more meaningful representation. The learning algorithm in S-PCA is very simple and the optimization procedure is robust and scalable to high-dimensional spaces.

EigenCuts is a graph-theoretic clustering algorithm based on a Markov-chain perspective. The key insight in this work is that bottlenecks between weakly coupled clusters can be identified by computing the sensitivity of the random walk to the variations in the edge weights. The EigenCuts algorithm performs clustering by removing bottlenecks identified by sensitivity analysis in an iterative fashion. As an efficient step in this process we also propose a specialized hierarchical eigensolver suitable for large stochastic matrices. For naturally-occurring datasets we demonstrate that the EigenCuts algorithm provides perceptually "meaningful" decompositions.

Joint work with Allan Jepson.

Applications such as news or video on demand, distance learning, scientific visualization, and immersive virtual reality must store, maintain, and retrieve large volumes of real-time data. The development of conventional real-time systems has mainly focused on supporting the real-time paradigm for computational and networking resources and usually assumes that all data resides in main memory. However, real-time multimedia systems manage large amounts of heterogeneous data (a heterogeneous mix of both real-time and non-real-time data), which require to be stored on secondary storage. In this talk, I focus on storage systems that can support the varied requirements of heterogeneous multimedia data. I will present our research in the areas of data placement, IO scheduling, and admission control for conventional disk-based storage systems. All our techniques are based on accurate extraction and modeling of hard disk performance parameters. This approach enables us to provide real-time guarantees without compromising system throughput. In the second part of the talk, I shall present our recent work on incorporating MEMS-based storage, an exciting new storage technology, into existing storage architectures and evaluate its application to streaming media servers. I will conclude by presenting an overview of SfinX, a high data-volume video surveillance application, that we are currently developing at UC Santa Barbara.

Biography

Raju Rangaswami is a doctoral candidate in Computer Science at the University of California, Santa Barbara. He received a B.Tech. degree in Computer Science from the Indian Institute of Technology, Kharagpur, in 1999. His primary research interest is building high-performance systems that are capable of offering quality of service (QoS) guarantees, with a focus on real-time storage systems.

Because of the increasing gap between processor frequency and Dynamic Random Access Memory (DRAM) speeds, one of the major architectural design considerations for any computer system is that of the memory subsystem. In most cases, the performance of the memory subsystem governs the performance of the system as a whole. This especially is true for modern, as well as future, symmetric multiprocessor (SMP) systems. Consequently, performance evaluation methodologies that facilitate the analysis and optimization of the memory subsystem are essential to the development of next-generation computer systems. In this talk, I will describe such a methodology, one that is based on sampled performance monitor event traces. The methodology will be described in the context of a performance evaluation study of sampled L2-cache data-load misses generated by the TPC-C benchmark executed on eight- and 32-processor of IBM's eServer pSeries 690 (p690) SMP systems. As will be demonstrated, sampled performance monitor event traces can be used to study memory performance, characterize process behavior, and improve application performance. In particular, they can be used to identify the resolution sites of level-two (L2) cache data-load misses and characterize the heavily-hit resolutions sites, i.e., level-three (L3) caches and main memory, with the goal of recognizing the heavily-hit regions of the application's address space, segments, pages, cache block, routines, instructions, and data structures. Information like this can be used to gain insights into application behavior on a multiprocessor to evaluate design aspects of and policies associated with the memory hierarchy with respect to workload demands.

Biography

Dr. Teller is an Associate Professor in the Department of Computer Science at the University of Texas at El Paso (UTEP). She received a Ph.D. in Computer Science from New York University in January 1991. Teller's dissertation focused on the design and performance evaluation of consistency-ensuring management of shared-memory multiprocessor translation-lookaside buffers (TLBs). Her research interests include parallel and distributed computing, computer architecture, performance evaluation and modeling, workload characterization, operating systems, and simulation methodologies. Teller's research has been funded by NSF, DARPA, the DoD, NASA, the U.S. Army Research Office, Sandia National Laboratories, Los Alamos National Laboratory, Intel Corporation, IBM, and Sun Microsystems.

For the past 18 months, I have been participating in a wide range of planning activities at DARPA, the Defense Advanced Research Projects Agency. In this talk, I will report on what I have learned during this time. This will include a mix of technical and non-technical topics. On the technical side, I will describe some of the current and future projects that DARPA is funding. On the non-technical side, I will compare the "venture capital" approach of DARPA to the "peer review" approach of NSF in terms of their effectiveness at advancing computer science research.

Biography

Dr. Dietterich (AB Oberlin College 1977; MS University of Illinois 1979; PhD Stanford University 1984) joined the Oregon State University faculty in January 1985. In 1987, he was named a Presidential Young Investigator for the NSF. In 1990, he published, with Dr. Jude Shavlik, the book entitled Readings in Machine Learning, and he also served as the Technical Program Co-Chair of the National Conference on Artificial Intelligence (AAAI-90). From 1992-1998 he held the position of Executive Editor of the journal Machine Learning. The American Association for Artificial Intelligence named him a Fellow in 1994, and the Association for Computing Machinery did the same in 2003. In 2000, he co-founded a new, free electronic journal: The Journal of Machine Learning Research. He served as Technical Program Chair of the Neural Information Processing Systems (NIPS) conference in 2000 and General Chair in 2001. He currently President of the International Machine Learning Society and he also serves on the Board of Trustees of the NIPS Foundation.

Society's major challenges continue to get larger, more complex, and more urgent -- all at increasing rates. Challenges which can only be dealt with collectively, where large groups of groups of people have to cooperate effectively.

Assumedly there's no question about the need to get collectively smarter. So ...? What are the prospects? What degree of improvement seems possible? Should we be in any hurry? How big a challenge would explicit pursuit represent? Are we on the path, or is haphazard wandering our current approach? Who should be responsible? Who could effectively participate?

Considering the number of people (indeed, of countries) needing to be closely involved with improving their capability to participate effectively in a new way of discovering and understanding applicable knowledge, and the rate of advancement in applicable technologies, and the new degrees of skill to be acquired, ... the scale of the challenge seems to be of unprecedented magnitude.

Thus, an effective strategical approach seems for sure to be necessary.

Strategy? Yes; and that's what I'll be describing. I've been working on it for decades now; I call it "The Bootstrap Strategy." And universities can begin to play an especially important role.

Biography

Douglas Engelbart, a 1948 OSU graduate in Electrical Engineering, has had a profound influence in the way that we use computers and their influence in our everyday lives. As early as 1962, he was involved in formulating a strategic framework by which computer systems could be integrated with one another and with millions of future users. The results of these strategies are employed today by all major computer manufacturers. Included in the results of this creativity are: the invention of the computer "mouse"; the development of the concept of "windows"; structure document files; hypermedia; integrated electronic mail. He is an OSU honorary Dr. of Engineering and a recipient of the American Ingenuity Award. He currently leads the Bootstrap Institute in Palo Alto, California.

Three exponentials have been the foundation of today's electronics, which are often taken for granted--namely transistor density, performance, and energy. Moore's Law captures the impact of these exponentials. Exponentially increasing transistor integration capacity, and exponentially increasing transistor performance, allowed us to realize complex architectures, delivering exponentially increasing performance. And exponentially decreasing energy per logic operation kept power dissipation within reasonable limits, making Moore's Law a reality. To keep this treadmill going, we will face exponentially increasing challenges, such as active and leakage power, variability, and design efficiency. These challenges, once addressed, will undoubtedly yield exponential rewards, as we have enjoyed in the past.

This talk will discuss potential solutions in all disciplines, such as microarchitecture, circuits, design technologies & methodologies, thermals, and power delivery. At ISSCC 2003, Gordon Moore showed that for every ant in the world today there are 100 transistors; our job is to grow that transistor number to 10,000. We'd like to see 100 times more transistors in the world for electronics and computing power by the end of the decade, and there is no reason we can't succeed.

Biography

Shekhar Borkar graduated with MS in Physics from University and Bombay, MSEE from University of Notre Dame in 1981, and joined Intel Corporation. He worked on the 8051 family of microcontrollers, the iWarp multicomputer project, and subsequently on Intel's supercomputers. He is an Intel Fellow and director of Circuit Research. His research interests are high performance and low power digital circuits, and high-speed signaling. Shekhar is an adjunct faculty member at Oregon Graduate Institute, and teaches VLSI design.

My talk covers an overview of implantable microelectronic systems and more specifically the development of an inductively-powered wireless microsystem for multichannel neural stimulation. This wireless stimulating microsystem is expected to have a variety of applications in implantable neural prostheses such as auditory, visual, spinal cord, and deep brain stimulators to restore peripheral and central nervous system abilities. A modular 1024-site microstimulator array with bidirectional wireless RF telemetry and 128 simultaneous stimulating sites, each capable of sourcing ±100mA is the ultimate goal. Some of the major challenges such as low power consumption, wideband data transmission, small size, large number of sites, and flexible stimulation strategies will be addressed.

The modular architecture of this wireless microsystem, where each microstimulator is a self-contained unit, allows achieving a large number of stimulating sites, while the system resources scale with the number of modules. A fully-integrated full-wave rectifier, capable of supplying up to 50mW with very small substrate leakage current, is used to reduce the implant size and power consumption. A phase-coherent FSK modulation protocol, a wideband inductive link, and a set of novel mixed-mode FSK demodulation circuits are employed to achieve the highest reported data rate of 2.5Mbps in an inductively powered device, using a 5/10MHz carrier. A closed-loop current source/sink is designed by utilizing MOS transistors in the triode region as linearized voltage-controlled resistors to realize compliance voltages up to 97% of the supply level, while showing output impedances in the excess of 100MÙ.

Biography

Maysam Ghovanloo received the B.Sc. degree in electrical engineering from the University of Tehran among the top-ten in the class of 1994. His undergraduate research was focused on developing an 8kW power supply for Nd-YAG Lasers. At the Etrat institute of technology, he worked on computer interfaces for industrial automotive robotic applications. He received the M.Sc. (Hons) degree in biomedical engineering from the Tehran Polytechnic in 1997. His M.Sc. thesis was on development of a multi-site physiologic recording system for investigation of the neural assemblies. From 1994 to 1998 he worked part-time at the IDEA Inc. (Tehran-Iran), where he participated in the development of the first modular patient care monitoring system in Iran. In December 1998 he founded Sabz-Negar Rayaneh Co. Ltd. to manufacture physiology and pharmacology research laboratory instruments such as precision bio-amplifiers, isolated stimulators, and data acquisition systems. He started the Ph.D. at the EECS department of the University of Michigan in Jan. 2000, majoring in circuits and microsystems. In summer-2002 he was with the Advanced Bionics Inc. working on the spinal-cord stimulator project. His Ph.D. research is on developing a wireless microsystem for neural stimulating microprobes. He has received awards in the operational category of the 40th and 41st DAC/ISSCC student design contest. He is a member of Tau Beta Pi, the engineering honor society, as well as the IEEE solid-state circuits and biomedical engineering societies.

Cryptographic engineering deals with hardware and software realizations of cryptographic functions, such as deterministic and true random number generators, secret-key cryptosystems (DES, AES, RC4, etc.), message-digest and authentication functions (MD5, SHA-X, HMAC, etc.), and public-key cryptosystems (RSA, Diffie-Hellman, Digital Signature Standard, and Elliptic Curve Cryptography). Of particular interest to the research community are implementations on embedded systems, since there are some significant challenges in creating efficient cryptographic implementations on embedded processors, due to low computing power, low dynamic memory for data, and limitations on program code size and stack memory usage. We will give an overview of the subject and summarize our research and development projects in cryptographic engineering, for example, efficient implementations of elliptic curve cryptography on embedded software platforms, creation of scalable and unified hardware arhcitectures for public-key cryptosystems, and the design of true random number generators.

Biography

Cetin Kaya Koc is a Professor in the School of Electrical Engineering and Computer Science at Oregon State University, where he joined in 1992.

He received B.S. (1980, summa cum laude) and M.S. (1982) degrees Electrical Engineering from Istanbul Technical University, and M.S. (1985) and Ph.D. (1988) degrees in Electrical and Computer Engineering from University of California, Santa Barbara.

He joined the Department of Electrical and Computer Engineering at the University of Houston in 1988 as an Assistant Professor, and then moved to Oregon State University in 1992, where he has been Assistant Professor (1992-1995), Associate Professor (1995-1998), Full Professor (since 1998), and Honors College Faculty (since 2003).

Prof. Koc is the founder and director of the Information Security Laboratory at Oregon State University. He received the OSU College of Engineering Research Award for Outstanding and Sustained Research Leadership in September 2001.

Prof. Koc's research interests are in cryptographic engineering, algorithms and architectures for cryptography, computer arithmetic and finite fields, parallel algebraic and numerical computation, and network security. He has founded the Workshop on Cryptographic Hardware and Embedded Systems (CHES) in 1999 and has been the program chair and proceedings editor from 1999 to 2003. He is currently serving in the steering committee of CHES. The Proceedings of CHES Workshops are published by Springer Verlag in the Lecture Notes in Computer Science (LNCS) series.

Prof. Koc was the Guest Editor of the special issue in April 2003 of IEEE Transactions on Computers on cryptographic hardware and embedded software development. He is also in the editorial boards of IEEE Transactions on Computers and IEEE Transactions on Mobile Computing.

Prof. Koc has been working as a consulting engineer with research and development interests in cryptographic design and high-speed computing for several organizations and companies including Pacific Northwest National Laboratories, Intel Corporation, RSA Security, and Samsung.

Prof. Koc is a senior member of IEEE and IEEE Computer Society.

Image segmentation is essential in many computer vision and image understanding applications. In our current work, we use a Bayesian network for object boundary detection in which the MPE (most probable explanation) before any evidence can produce multiple non-overlapping, non-selfintersecting closed contours and the MPE with evidence--where one or more connected boundary points are provided--produces a single non-self-intersecting, closed contour that accurately defines an object s boundary.

Further, we present a linear-time algorithm that determines the MPE by computing the minimum-path spanning tree of a weighted, planar graph and finding the excluded edge that forms the most probable loop. This allows for real-time feedback within an interactive environment in which every mouse movement produces a recomputation of the MPE based on the new evidence and displays the corresponding closed loop. We call this interface "object highlighting" since object boundaries appear and disappear as the mouse cursor moves.

Biography

Professor Eric N. Mortensen graduated Magna Cum Laude with a B.S. in computer science in 1991 from Brigham Young University, an M.S. in computer science in 1995, and a Ph.D. in computer science in 2000, also from Brigham Young University. He has been an assistant professor at Oregon State University since August 2001.

Dr. Mortensen's research interests include image processing, computer vision, image and video segmentation and editing, and image-based modeling and rendering. He has authored or co-authored several papers on semi-automatic image segmentation, image editing, image-space contour interpolation, and color quantization.

Graph-based approaches for sequencing motion capture data have resulted in the generation of some of the most realistic and controllable character motion to date. Previous graph-based approaches have employed a run-time global search to find paths through the motion graph that met user defined constraints such as a desired locomotion path. Such searches do not scale well to large numbers of characters. In this talk, I will describe an approach that benefits from the believable motion of graph-based approaches while maintaining basic user control and scaling well to large numbers of characters. Our approach is based on the precomputation of an all pairs shortest path matrix built directly from a pose transition graph and on a mobility map that encodes the movement possibilities for a character from a given pose. We combine these two structures and at run time, we perform a local search to generate real-time natural motion sequences that achieve user constraints. We demonstrate the quality of the motion through various locomotion examples including target tracking and collision avoidance. We demonstrate scalability by computing motion for crowds of up to sixty articulated walking characters at real-time rates.

Biography

Asst. Prof. Ron Metoyer joined the computer science department at Oregon State University in the fall of 2001 with a Ph.D. from the Georgia Institute of Technology where he worked in the Graphics, Visualization and Usability Center. His dissertation research involved behavioral modeling, specifically, modeling and visualizing the motion of pedestrians in urban and architectural scenes. Currently, Dr. Metoyer leads the Interactive Graphics and Vision Lab (IGVL) along with his colleague, Dr. Eric Mortensen. He and his students conduct research in creating interactive spaces for training and education. Research projects range from the development of believable character motion generation algorithms to the development of interfaces for novice users to create and interact with 3D content. One testbed application is a 3D immersive quarterback training environment for coaches and quarterbacks to use as a training and visualization tool. This work was recently funded by an NSF CAREER award entitled, "Understanding the Complexities of Animated Content".

In this talk, I will describe how we make use of the cognitive dimensions framework to describe and measure API usability at Microsoft. The cognitive dimensions framework was originally proposed by Green et al at the University of Cambridge to describe programming language usability. We have made some small adaptations to the framework so that it is better suited to discussions of API usability. I will describe these adaptations, our experiences in using the framework at Microsoft and the significant advantages that the framework has provided us in being able to understand what makes an API usable.

Biography

After completing my PhD in Computing Science at the University of Glasgow, Scotland, I worked as a developer for two years at Motorola in Scotland, using Visual C++ to build smartcard development tools. After Motorola, I moved to Seattle to join the Visual Studio team at Microsoft, working as a usability engineer. My main responsibilities are working with API teams throughout Microsoft to ensure that they have the information they need in order to design as usable a set of APIs as possible.

High speed optical and electrical links are being transformd from simple binary transceivers to sophisticated digital communication systems employing various combinations of modulation, coding, and adaptive filtering for equalization and interference cancellation. This transformation is predictable given the bandwidth limitations imposed by both optical fibres and chip-to-chip interconnects. However, the efficient implementation of these signal processing functions in integrated circuits provides exciting research opportunities.

This talk will provide an overview of the work being done at the University of Toronto in the area of high speed signaling. It will include brief overviews of several projects underway at U of T including integrated circuits for transmission, equalization, timing recovery, and error correction at multiple Gb/s. The presentation will also provide a detailed description of channel modeling methods for behavioral simulation, and a novel adaptation algorithm suited to high speed equalizers in the presense of jitter. The algorithm allows vertical eye opening to be traded for increased timing margin at the receiver.

Biography

Tony Chan Carusone completed a B.A.Sc. at the University of Toronto, Engineering Science Division in 1997 where he received the Governor-General's Silver Medal awarded to the student graduating with the highest average from approved university programs. In 2002 he obtained a Ph.D. in the Department of Electrical and Computer Engineering, also at U of T. Since 2001, he has been an Assistant Professor in the ECE department at U of T where he is a member of the electronics group and the Information Systems Laboratory. In 2002 he was awarded the Canada Research Chair in Integrated Systems, and named an Ontario Distinguished Researcher. In addition, he has worked in industry doing both analog and digital design with large and small companies including, Celestica, Analog Devices, Snowbush, and Gennum. He is a member of the IEEE and has several publications in the areas of analog filters, adaptation, and chip-to-chip communication. His research interests are in integrated circuits for high speed signal processing, both analog and digital.

Teachers may engage in end-user programming to support student learning or administrative activities associated with teaching. The objective of this research is to understand strategies used by teachers in program comprehension and to identify specific problems they face. A think-aloud study was conducted of teachers comprehending an event-driven application, consisting of a graphical user interface and the scripts controlling it. We found that end users followed a strongly top-down strategy and breadth-wise exploration of the application. Depth-wise exploration was observed in half the teachers. Teachers varied greatly in their motivations and persistence to dig deeply into the code. Problems of the teachers included difficulties comprehending the event-driven application, given the distributed nature of the code, choosing appropriate inputs for running the program, and reasoning about the results of their test runs.

Biography

Dr. Wiedenbeck teaches in the area of human-computer interaction. Her research interests include human-computer interaction, empirical studies of end-user programming, interface design and evaluation, computer-supported cooperative work, computer training and individual factors affecting the learning and use of IT.

One of the great pleasures about working in the field of scientific and engineering visualization is that there are no rules. Anything that turns data into information-rich understanding is fair game. Thus, visualization has become a big bag of tricks. As such, much of the effectiveness of visualization comes in matching the right trick to the right dataset. And, the great thing is, there are new tricks coming all the time! This talk will discuss four of those new tricks, and the types of data that they allow us to better understand:

1. Desktop interactive volume rendering

OSU's Volume Explorer program (vx) allows interactive direct volume manipulation on the desktop. It can do this with datasets as large as 1 billion voxels. This has allowed us to visualize many 3D datasets that we would otherwise not have touched. This includes new medical and fluids data, as well as a few non-traditional volumes such as mechanical systems extruded in time, geophysics, and a Confederate submarine.

2. Color 3D models

OSU's 3D Visualization Hardcopy project is going to be instrumental in creating interesting collaborations both within and outside of EECS. There seems to be no end to the useful 3D models that can be created that act as visualization aids, especially when the models are in color. This portion of the talk will discuss some of the most recent 3D Hardcopy projects which include human body parts for doctors to experiment on, 3D ultrasounds, planets, and even a 3D model of the smoke from last year's Southern California fires.

3. Visualization using Graphics Hardware Shaders

Graphics Hardware Shaders are one of the most exciting things to happen to computer graphics in years. Hardware Shaders allow you to write your own graphics code and have it executed on the graphics board as part of the display process. This portion of the talk will discuss what this is and what it means to futuristic visualization and game programming.

4. True 3D for Visualization

Normal graphics display devices create the illusion of 3D on a flat monitor or projection screen. True 3D displays are devices that draw 3D objects in where they really belong in actual 3D space. We have been using one of these devices, and have found that they work best for sparse visualization applications such as applications in fluid flow and molecular modeling.

Biography

Mike Bailey is a new CS faculty member, having just arrived from the University of California San Diego. Mike received his PhD from Purdue University. His area of interest is scientific and engineering visualization, including some of the new and innovative methods described in this talk. He is teaching CS 519, Scientific Visualization, in the Fall Quarter.

In this talk I'll describe my interests and recent work in the area of machine learning for structured data. Example problems include learning to interpret hand-written equations, hand-written sketches, and noisy video sequences. Humans are remarkably robust at forming such interpretations, while computers remain far behind. Accordingly, machine-learning researchers are now pursuing new models and learning algorithms for structured data types.

My recent work in this area is focused on interpreting relational sequences. I'll describe the "simple-transition cost model" for relational sequence data. The model is parameterized by a set of weighted logical constraints. Importantly, inference in this model, i.e. finding a minimum-cost interpretation, has a practically efficient decomposition based on logical reasoning and bounded search. I'll describe how to learn such a model, and then present promising experiments in relational video interpretation.

Biography

Alan Fern is a new faculty member in Computer Science at Oregon State University. In 2004 he received a Ph.D degree in Computer Engineering from Purdue University. He received an M.S degree from Purdue in 2000 and a B.S from the University of Maine in 1997. His primary research interests are in machine learning, data mining, and automated planning/control.

Recent dramatic improvements in the accuracy of automatic transcription for spontaneous conversational speech hold the promise to unlock access to vast quantities of spoken language sources. Realizing that promise requires that we develop search technology that is tuned to the nature of conversational speech. In this talk, I will describe a first effort to do just that, leveraging a large collection of "oral history" interviews for which a uniquely rich collection of associated metadata is available. I'll briefly describe the status of our work on speech recognition, topic segmentation, and text classification. I'll then focus on the process that we have used to build an information retrieval test collection, and our results from initial experiments with that collection. I'll conclude by explaining how those results are helping to guide future work on speech recognition, and our plans for building test collections for languages other than English. This is joint work with Charles University, the IBM TJ Watson Research Center, the Johns Hopkins University, the Survivors of the Shoah Visual History Foundation, and the University of West Bohemia.

Biography

Douglas Oard is an Associate Professor at the University of Maryland, College Park, with a joint appointment in the College of Information Studies and the Institute for Advanced Computer Studies. He holds a Ph.D. in Electrical Engineering from the University of Maryland, and his research interests center around the use of emerging technologies to support information seeking by end users. Dr. Oard's recent work has focused on cross-language information retrieval, searching spoken language collections, data mining from text, and the exchange of ratings by networked users. Additional information is available at http://www.glue.umd.edu/~oard/.

This talk will go into the current status of the 2.6 Linux kernel development process, detailing the prior history of the 2.5 development process, and attempt to explain the brash claim by the kernel developers that a 2.7 fork will not happen any time soon as they "are happy the way things currently are." It will also go into how the development process for the kernel maintainers has changed without anyone else noticing, and how the whole kernel development process seems to be breaking all traditional software engineering rules about how a project should be successfully run.

Biography

Greg Kroah-Hartman is currently the Linux kernel maintainer of USB, PCI, and a handful of other subsystems. He is also in charge of the linux-hotplug and udev projects, along with maintaining the Gentoo Linux kernel package. He is a contributing editor to Linux Journal, writing a bi-monthly column about Linux kernel driver programming, and is the co-author of the upcoming "Linux Device Drivers, third edition" book from O'Reilly. He works for IBM in Portland, Oregon, doing various Linux kernel related things.

The Evolutionary Requirements Analyser (ERA) applies evolutionary computing techniques to automatically select optimal combinations of human and machine agents in a system model to match non-functional requirements (NFRs). The tool assesses the reliability, performance times and cost of different system models by executing many model variants, as evolving forms, with scenarios and different combinations of environmental variables. Better performing models are selected, to converge on an optimal solution. Use of the tool is illustrated with a case study of requirements analysis for component selection in a command and control system.

Biography

Alistair Sutcliffe is a full Professor of Systems Engineering, and Director of the Centre for HCI Design, in the School of Informatics, University of Manchester, UK. His research spans software engineering, human computer interaction and cognitive science, with current interests in scenario based design, methods for requirements engineering, and creative design for the Internet. He is a leading authority on multimedia user interfaces, has authored 6 books and 200+ publications on user interface design, requirements engineering, software and domain knowledge reuse. He serves on the editorial boards of several journals in the software engineering and human computer interaction, and recently co-chaired the ACM conference Designing Interactive Systems 2002.

The current ``best-effort'' Internet does not guarantee Quality of Service (QoS) such as minimum bandwidth, packet loss rate, and delay which are critical to many multimedia applications. As such, many significant challenges remain to design and deploy delay sensitive multimedia applications over the Internet effectively. In this talk, I will present the path diversity (PD) framework for concurrent media streaming to a receiver using multiple routes. Without requiring QoS, the PD framework improves the quality of the streamed media via multiple routes created using either multiple senders or relay nodes, in order to increase available bandwidth, reduce packet loss and delay. I will present source and channel coding techniques together with transport protocol in the PD framework to improve the quality of streamed media.

Biography

Thinh Nguyen is currently assistant professor in the school of electrical engineering and computer science at Oregon State University. He received his Ph.D and MS from UC Berkeley, and BS from University of Washington. Before joing OSU, he worked for a number of companies and research lab including Intel, Microsoft, Siemens, and Lawrenrence Livermore National Lab.

Convergence is a central problem in both computer science and in population biology. Will a program terminate? Will a population go to an equilibrium? In general these questions are quite difficult – even unsolvable. In this talk we will concentrate on a very simple iterations of the form

    x_{t+1} = f(x_t)

where each xt is simply a real number and f(x) is a reasonable real function. Such one-dimensional iterations have special properties not shared by higher dimensional iterations. For example, if f(x) is continuous and bounded, with exactly one fixed point and no cycle of period 2, then the iteration will always approach the fixed point. For such a system, we say that an iteration is “globally stable” if it approaches the fixed point for all starting points, and that the iteration is “locally stable” if it approaches the fixed point for all initial points near enough to the fixed point. Local stability can be checked by a simple derivative test, but showing global stability requires more effort. Population biologists have used one-dimensional iterations as simple models of population growth. When they have shown local stability, they have assumed global stability and have never run into any difficulty. Over the years a variety of mathematical techniques have been used to show that specific simple population models exhibit this local stability implies global stability phenomenon. But, no single method could prove this result for all the commonly used population models.

We will show that there is a simple method which assures global stability. Our method uses bounding of f(x) by a ratio of two linear functions. We call this bounding “enveloping” and we show that enveloping implies global stability. Returning to the population models, we show that for these models local stability implies enveloping and hence global stability. For several models a single enveloping function can be used for all parameter values, and we give a Taylor series condition which implies this enveloping. For other models, different enveloping functions are needed for different parameter combinations. Although enveloping is a sufficient condition for convergence, we show by example that enveloping is not necessary. Finally we show that enveloping implies global stability even when f(x) is a discontinuous multi-function.

Biography

Paul Cull is a professor of computer science at Oregon State University where he has worked for over 30 years. He teaches a variety of undergraduate courses and graduate courses in theory of computation, analysis of algorithms, and cybernetics. During the summers he works with students in a research experience for undergraduates program which focuses on problems at the interface between computer science and mathematics. In 2001, Prof. Cull was given the Alumni Professor Award.

Paul Cull's research interests are mathematical biology, theory of computation, and analysis of algorithms. His early work was on inferring gene linkage from pedigree data. He earned his Ph.D. at the University of Chicago where he studied with the committees on mathematical biology and information science. He wrote his thesis on the analysis of neural networks.

Over the years, he has worked on a large variety of topics and published papers in many venues. His most recent papers are in machine learning, neural nets, interconnection networks, biological string alignment, discrete iterations, and error-correcting codes. He is completing a book on difference equations.

The confluence of virtual reality and artificial life, an emerging discipline that spans the computational and biological sciences, has yielded synthetic worlds inhabited by realistic, artificial flora and fauna. Artificial animals are complex synthetic organisms that possess functional biomechanical bodies, perceptual sensors, and brains with locomotion, perception, behavior, learning, and cognition centers. Artificial humans and other animals are of interest in computer graphics because they are self-animating graphical characters that can dramatically advance the state of the art of production animation and interactive game technologies. More broadly, these biomimetic autonomous agents in their realistic virtual worlds also foster deeper, computationally oriented insights into living systems.

Biography

Demetri Terzopoulos holds the Lucy and Henry Moses Professorship in Science at New York University and is Professor of Computer Science and Mathematics at NYU's Courant Institute. He is also affiliated with the University of Toronto, where he is Professor of Computer Science and Professor of Electrical and Computer Engineering. He graduated from McGill University and received the PhD degree (EECS) from MIT in 1984. Prior to becoming an academic in 1989, he was a program leader in Schlumberger corporate research labs in California and Texas. His published work comprises hundreds of research papers and several volumes, primarily in computer graphics, computer vision, medical imaging, computer-aided design, artificial intelligence, and artificial life. He has given hundreds of invited talks around the world, including numerous distinguished lectures and keynote addresses. A Fellow of the IEEE, Professor Terzopoulos has been a Killam Fellow of the Canada Council for the Arts, a Steacie Fellow of the Natural Sciences and Engineering Research Council of Canada, and a Fellow of the Canadian Institute for Advanced Research. At present, he is a Senior Fellow at UCLA's Institute for Pure and Applied Mathematics (IPAM). His many honors include computer graphics awards from Ars Electronica, NICOGRAPH, Computers and Graphics, and the International Digital Media Foundation, and computer vision awards from the IEEE, the Canadian Image Processing & Pattern Recognition Society, the American Association for Artificial Intelligence, and the International Medical Informatics Association. Terzopoulos is the conference co-chair of the 2005 ACM/Eurographics Symposium on Computer Animation (SCA) and was program co-chair of Pacific Graphics 2004 and the 1998 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). He has been a founding editorial board member of four journals spanning the vision, graphics, and medical imaging fields.

The talk will describe the structure of the reorganized CISE and the role of the Computer Communications Foundations (CCF) Division. Cross-cutting initiatives and High-End Computing (HEC) programs and themes will be discussed, along with the trends in computer architecture funding.

Although delta-sigma converters reduce the effects of analog element imperfections on the accuracy and linearity of conversion, they suffer from residual problems which cannot be solved by any combination of oversampling and error shaping. These include path mismatch errors in cascade multi-stage converters, and multi-bit DAC nonlinearity errors for wideband ADCs.

In this lecture, several digital techniques will be described which can be used to acquire and suppress such errors. They are typically based on correlation between a known signal (such as a dither or element selection data) and the unknown error. Their operation is not dependent on oversampling, and hence remains effective for wideband converters. Several examples will be shown to verify the effectiveness of digital correction methods.

Biography

Gabor Temes, Professor, Electrical and Computer Engineering Department, Oregon State University, Professor Emeritus, UCLA. Formerly with UCLA, Ampex Corp., Stanford University and BNR. Life Fellow, IEEE. He received the Technical Achievement Award and the Education Award of the IEEE CAS Society, as well as the IEEE Centennial Medal. He has written many books and papers on discrete and integrated circuit design. He is the recipient of the 1998 IEEE Graduate Teaching Award and received the IEEE Millennium Medal and the IEEE-CAS Golden Jubilee Medal in 2000.

Biometrics holds much promise in a wide variety of contexts to help increase security. Research at the moment is mainly focused on the technical challenge of developing these systems, e.g. reliable methods of matching based on biometric characteristics. This talk provides examples of current and future applications of biometrics. Concepts that need to be considered when designing and/or evaluating biometric systems from a HCI perspective are introduced. Some indicators on where and why current systems fail are discussed.

Biography

Dr. Simone Stumpf just joined Oregon State University from University College London (UCL), where she has been working on designing knowledge management systems for retail security personnel and, more recently, evaluating the usability of biometric systems. She received the B.Sc. degree from UCL in 1996 and the Ph.D. from UCL in 2001.

Miniaturized bio / chemical analysis system (Lab-on-a-Chip, ƒÝTAS) has been of great interest over the past decade. By precisely controlling small amounts of biological or chemical samples, fast and accurate analysis can be achieved on a compact system with low cost. This talk will briefly review the motivation and advantages of such systems and provide examples of developed systems in life science.

Biography

Arum Han is currently a Ph. D Candidate in the School of Electrical & Computer Engineering at the Georgia Institute of Technology. He received the B.S. from Seoul National University in 1997 and the M.S. from University of Cincinnati in 2000.

The task of visual motion detection is to visually identify moving objects, extract information about their shapes, locations, speed, directions, and so on. It is very useful in many applications such as robot motion control, video compression, navigation control for vehicles and aircrafts, and high speed motion analysis. Traditional solution is to use digital camera plus digital processor system, which normally leads to intensive computation load and data transfer load. The new solution is to use smart visual motion sensors as front-end, which extracts motion information at pixel level and transfer only the extracted data to digital processor, thus, largely reduces the computation and data transfer load. However, previous visual motion sensors suffer from large pixel size, high power consumption, low accuracy, etc. To deal with these issues, a novel time stamped structure for 2D visual motion detection is proposed. It captures motion in more than 100 times higher time resolution than digital cameras with the same frame rate. Several CMOS IC chips have been fabricated which proved the feasibility and also optimized for pixel size, power consumption, accuracy, etc. The final version achieves 18 times smaller pixel size, 5000 times lower pixel power consumption compared with most recent related works.

Furthermore, the time stamped structure has been successfully expanded to digital imager design, which largely extended the dynamic range of the imager for both strong and weak light. By automatically adjusting integration time of each pixel, 120dB dynamic range has been achieved, which is 1000 times wider than most commercial products.

Several other research projects will also be briefly talked about, including self-calibrated transmitter driver, laser radar speed/range finder, ultra-low power imager for wireless network, and modeling of vertical-cavity surface-emitting laser diode.

Biography

Guangbin Zhang just received his Ph.D. degree in the Erik Jonsson School of Engineering and Computer Science of the University of Texas at Dallas (UTD) in Dec. 2004. Before that, he received his Bachelor's Degree from the Automation Department of Tsinghua University in July, 1997 and his Master's degree from the Semiconductor Institute of the Chinese Academy of Sciences (CAS) in July, 2000. His research interests focus on analog and mixed signal integrated circuit design, especially on smart imager sensors and high speed data transmissions. He is also a student member of the IEEE Solid-State Circuits Society.

The need for storing information goes as far back as recorded history itself. Today, hard disk drives serve this need as arguably the most sophisticated mass storage devices ever. However, with areal densities (in Gbits/in2) road-mapped to grow ten-fold every decade, the shrinking magnetic ‘bit’ is threatened by superparamagnetism or the spontaneous erosion of its magnetization over time.

In this magnetically-stored presentation, I will introduce the principle of magnetic recording and discuss my work on superparamagnetism and the challenges in engineering an antidote to it.

Biography

Dr. Pallavi Dhagat holds a doctoral degree from Washington University, St. Louis, where her research focused on the characterization of thermal reversal of magnetic grains. She has an extensive background in magnetic recording technology as a recording physicist.

Prior to joining OSU as an instructor, she was a research engineer in the Advanced Concepts Group at Seagate Technology, Minneapolis. Here she made significant contribution to the development of perpendicular recording technology.

Her research interests include information storage, magnetization dynamics, magnetic nano-metrology, bio-magnetic sensors and distributed sensor systems.

This talk focuses on the finite-terminal random multiple access in wireless networks. We characterize the relations among the throughput region of random multiple access, the capacity region of multiple access without code synchronization, and the stability region of ALOHA protocol, all over a standard multi-packet reception (MPR) channel. We first show that if the MPR channel is standard, the throughput region is co-ordinate convex. We then study the information capacity region of multiple access without code synchronization, and show that the asymptotic capacity region equals the throughput region. Next, we study the stability region of ALOHA protocol. For a class of packet arrival distributions, we show that the stationary distribution of the queues possesses the positive and strong positive correlation properties; and this consequently gives an outer bound to the stability region. We also show that if a conjectured “sensitivity monotonicity” property can be shown for the stationary distribution of the queues, then the equivalence between the closure of the stability region and the throughput region follows as a direct consequence, irrespective of the packet arrival distributions.

Biography

J. Rockey Luo received the B.S. and M.S. degrees in Electrical Engineering from Fudan University, Shanghai, P.R.China, in 1995 and 1998, respectively. He received the Ph.D. degree in Electrical and Computer Engineering from University of Connecticut in 2002. Since September 2002, he has been a Research Associate with the Institute for Systems Research (ISR), University of Maryland, College Park. His research interests are in wireless networks, wireless communications, energy issues, information theory and signal processing.

http://www.isr.umd.edu/~rockey

In recent years, the demand for power sensitive designs has grown significantly due to the fast growth of battery-operated portable computing devices. In this talk it will be shown that sub-threshold leakage current can be used for ultra-low power computation in applications such as cell phones, PDA, pacemakers etc. It will also be shown that an integrated device/circuit/architecture co-design approach will significantly improve the performance of digital sub-threshold operation. Result on a five-tap FIR filter shows 100MHz performance with ~100X less power than the normal operation.

Biography

Bipul C. Paul received the B. Tech. and M. Tech. Degrees in radiophysics and electronics from the University of Calcutta, India, and the Ph.D. degree from the Indian Institute of Science, Bangalore, India. He was with Alliance Semiconductor, where he led the project on synchronous DRAM design. He is presently working as a Post Doctoral Fellow at the School of Electrical and Computer Engineering, Purdue University, West Lafayette, IN. His present research interests include low-power design of nanoscale circuits under process variation (using both bulk and SOI devices), testing, and noise analysis. He is also working on the device and circuit optimization for ultra-low power digital sub-threshold operation.

Dr. Paul received the Best Thesis of the Year award in 1999.

Sigma-Delta modulation is the modern technique to perform high resolution analog-to-digital conversion. A Sigma-Delta modulator (SDM) consists of a feedback system that coarsely quantizes an oversampled analog signal. From a signal processing point of view, an SDM is a system of high simplicity, but paradoxically of difficult rigorous analysis. Due to the presence of a nonlinear operation (quantization) in the feedback loop, one is not able to derive the explicit expression of the output of an SDM in time. With the classic white-noise assumption of the quantizer error, one "linearizes" the quantization operation, and thus facilitates the analysis, but one only obtains partial information on the error signal with limited validity. A genuine explicit expression of the output is necessary to study refine properties of the error such as correlation with the input, idle tones, etc...

In this talk, we show for a generic family of Nth order SDM's (among other conditions, the N zeros of the noise transfer function must be located at z=1) with constant inputs, that the system is mathematically equivalent to a linear feedback system, followed by an N-dimensional nonlinear but instantaneous operation. In other words, the nonlinear mechanisms of the SDM can be mathematically extracted from the feedback loop for signal analysis. The explicit derivation of the equivalent instantaneous nonlinear operation is still difficult, but its existence is mathematically proved, and an approximate knowledge of this function is already sufficient to derive error properties of SDM's not predictable by white-noise modeling. We finally explain how this theory can be extended to time-varying inputs.

Biography

Nguyen T. Thao received his Ph.D. degree in Electrical Engineering from Columbia University, New York, in 1993. He is currently an Associate Professor in the Department of Electrical Engineering of the City University of New York. His current research emphasis is the theoretical signal analysis of Sigma-Delta modulation.

We present a novel test data compression architecture and associated design techniques capable of achieving almost exponential reduction in test data volume and test time while allowing use of commercial Automatic Test Pattern Generation (ATPG) tools. Over the past 20 years, the presence of unknown logic values or X's in industrial designs has been widely recognized as the most significant barrier to efficient implementation of test response compaction that is necessary for Built-In Self Test and test compression applications. X-Compact is an new test response compaction technique that overcomes this barrier by tolerating X's while achieving close to exponential reduction in the test response data volume, without compromising test quality and diagnosis capability for most practical scenarios. Xpand enables close to exponential reduction in test stimulus data volume without requiring any significant changes to current design and test flows. The architecture has been implemented in more than 30 industrial designs.

Biography

Subhasish Mitra is a Senior Staff Engineer at Intel Corporation, a Consulting Assistant Professor in the Electrical Engineering Department of Stanford University, and the Associate Director of the Center for Reliable Computing of Stanford University. His research interests include robust computing, VLSI design and test, and computer architecture. He received Ph.D. in Electrical Engineering from Stanford University in 2000.

At Intel, Dr. Mitra is responsible for developing enabling technologies for Design for Excellence (DFX) - Design for Testability, Reliability, Manufacturability and Debug - in advanced process technologies. At the Center for Reliable Computing (CRC) of Stanford University, he supervises Ph.D. students and is currently involved with the Stanford CRC test chip experiment projects. During 2000-2001, he provided consulting at Agilent Technologies in their System Chip Testing program.

Dr. Mitra has published more than 60 technical papers in leading conferences and journals, and invented design and test techniques that have seen wide-spread proliferation in the industry. His most recent award is the Intel Achievement Award, Intel's highest corporate award, that he received in April 2004 "for the development and deployment of a breakthrough test compression technology." He is actively involved with the organization of several IEEE-sponsored conferences.

Many work practices consist of repeated discussions among groups of people. For example, software developers talk to customers to generate requirements, brainstorm and sketch design alternatives and make decisions, and perform code walkthroughs and reviews. A large amount of the rich, unstructured information that is generated during these discussions often does not get recorded as formal knowledge. Yet this information is later useful for providing additional context, details, and decisions surrounding a project.

In this talk, I will present several prototypes that demonstrate the application of ubiquitous computing to automatically capture, integrate and store multimedia records of different work discussions. We capture knowledge acquisition sessions with Tagger, and have evaluated the use of captured sessions in creating a requirements document. With the TeamSpace system, we have achieved authentic, long-term use of capture and access in general meetings. I will discuss what we have learned about the motivations and potential benefits for using recorded meeting information. I will also discuss the behavioral patterns we observed and the implications for supporting users with meeting capture and access systems.

Biography

Heather Richter is a Ph.D. candidate in the College of Computing at the Georgia Institute of Technology. She is a member of the Ubicomp Research Group and GVU Center, where she investigates the use of meeting capture and access systems. Her research interests include Human-Computer Interaction, Software Engineering, and Ubiquitous Computing. She received her B.S. in computer science from Michigan State University in 1995.

Vast amounts of scientific data are collected everyday. Tools for automatically extracting interesting patterns from the data with limited or no human supervision are critical toward understanding and gaining knowledge from the data. In this talk, I will present my research on two unsupervised pattern discovery problems within the context of Earth Science applications.

The first problem, clustering multi-spectral land cover data, poses fundamental challenges to existing unsupervised learning techniques due to the high dimensionality of the data. I will present a method based on random projection and cluster ensemble techniques to tackle the issues caused by high dimensionality. Results on land cover data and other benchmarks will be presented to show that this approach significantly improves the clustering performance on high dimensional data over existing approaches.

In the second problem, correlation pattern analysis of vegetation-precipitation data, I will build on the classical statistical method of Canonical Correlation Analysis (CCA) and introduce a novel approach to learning mixtures of CCA models. This approach not only addresses the linearity limitation of the traditional CCA method, but also is capable of finding correlation patterns that are only locally valid in the data. I will demonstrate the efficacy of the proposed approach with both synthetic and real-world data.

Biography

Xiaoli Fern is a Ph.D. candidate in Computer Engineering at Purdue University. She received her B.S. and M.S. degrees from Shanghai Jiao Tong University in 2000 and 1997 respectively. Her primary research interests are in machine learning and data mining. She is particularly interested in developing practical algorithms for automatically discovering useful patterns from scientific data.

Knowledge work is intensely collaborative, and computers have an increasing role in mediating collaboration. This talk covers three projects on collaborative technologies at the Berkeley Institute of design. The first is "Multiview" which is the first practical video-conferencing system to provide spatial faithfulness between *groups* of participants working together. In particular it is the first to preserve gaze and gesture cues between two collaborating groups.

The second project is "Livenotes" which supports live, collaborative note-taking in *small* groups of students in a live lecture. Livenotes seeks to re-create the a small- class experience for students in very large classes. Both Multiview and Livenotes will appear at the CHI conference here in April. The third project is "ABC" or Activity- Based Computing. ABC is really a meta-project with many application threads. We start with fine-grained user log data from normal desktop and laptop computer use. From this we extract patterns of "Activity" which include the various ad-hoc projects that people work on. We believe that this activity information is a large part of the understood "context" between people that computers have traditionally lacked. The applications we are working on include pro-active document sharing, access control, pro-active retrieval, command predication and disambiguation, and attention management.

I'll describe some early results on the public Enron email dataset. I will briefly describe our work on cryptographic methods to protect the privacy of user data even as it is being mined.

Biography

John Canny is the Paul and Stacy Jacobs Distinguished Professor of Engineering. His research is in human-computer interaction, with an emphasis on modeling methods (usually probabilistic) and privacy approaches using cryptography. He received his Ph.D. in 1987 at the MIT AI Lab. His dissertation on Robot Motion Planning received the ACM dissertation award. He received a Packard Foundation Faculty Fellowship and a Presidential Young Investigator Award. His peer-reviewed publications span robotics, computational geometry, physical simulation, computational algebra, theory and algorithms, information retrieval, HCI and CSCW and cryptography. He has best-paper prizes in IEEE FOCS, ECAI (European AI Conference) and AAAI.

Over the past several decades, the compiler research community has developed a number of sophisticated and powerful algorithms for a variety of code improvements. Although there are still promising directions for particular code optimizations, programming languages, and machine architectures, research on optimizations is nearing the point of diminishing returns and other approaches are needed to achieve further performance improvements. Our research aims to address this challenge by investigating and developing an innovative framework and system for continuously and adaptively applying code improvements. Our system, the Continuous Compiler (CoCo), determines "optimization plans" at compile-time that describe the best way in which to apply both static and dynamic code transformations. The plans consider program and machine context, interaction among optimizations, and performance profit. Through such planning, CoCo can tailor and adapt its decisions to more synergistically apply a whole suite of code transformations. This talk will describe the Continuous Compilation approach and present initial results. These results include novel analytic models that can accurately predict the performance benefit (profit) of applying an optimization without actually doing it or running the resulting program code. Using the analytic models, we have developed several planners that can guide a static optimizer. Initial results are very encouraging and show that optimizations can be effectively directed by planning. The talk will conclude with a brief discussion of CoCo's run-time system, including its dynamic code translator, instrumentation optimizer, and source-level debugger for dynamically translated code.

Biography

Bruce Childers is an Assistant Professor in the Department of Computer Science at the University of Pittsburgh. Dr. Childers received a BS degree (Computer Science, 1991) from the College of William and Mary, and a PhD degree (Computer Science, 2000) from the University of Virginia. His research interests are compilers and software development tools, computer architecture, and embedded systems. Current projects include continuous compilation to synergistically apply both static and dynamic code optimizations, debugging for dynamically translated code, power-aware real-time systems, and demand-driven software testing.

Computer privacy, especially online, is an increasingly important concern for users, and a difficult challenge for businesses facing increasing regulatory and public scrutiny. System designers face the difficult task of not only ensuring that user information is adequately protected, but also that users can adequately understand and control their exposure to risks. To address these concerns, we need to learn how users think and reason about privacy, and include privacy as a requirement in design. In the past, researchers have advocated using guidelines and heuristics to both identify problems and suggest design solutions. I describe and discuss current design guidelines and heuristics, and the reasons why, despite great interest and need, they have failed to make an impact on today’s design practices. I will present STRAP, a novel analysis framework for the structured analysis of privacy in design. In STRAP, I combine elements of goal-oriented analysis with heuristics to address many of the problems identified with previous frameworks. I show how this method can be used to perform a more thorough analysis and derive better design solutions. I will also present a number of studies I have conducted that examine how users think about privacy and the problems they face in managing risks, and how these conflict with current privacy practices and interface designs. I conclude by presenting a set of interface design guidelines for next-generation privacy awareness and management interfaces.

Biography

Carlos Jensen is a Ph.D. candidate at the College of Computing and GVU Center at the Georgia Institute of Technology, specializing in the study of Human-Computer Interaction and Privacy. The focus of his research is on understanding how system design affects decision making, especially in the areas of privacy, trust, and online communities. He is a member of the Personal Policy Lab at Georgia Tech, and The Privacy Place consortium.

Modeling and estimating software reliability during testing is useful in quantifying the quality and dependability of the developed software systems. However, such measurements applied so late in the development process leave done little to improve the quality of the software system in a cost-effective way. Reliability, an important dependability attribute, is defined as the probability that the system performs its intended functionality under specified design limits. We argue that in order to build reliable systems, reliability (and other dependability attributes) must be "built into" the software development process. Consequently, reliability models must be adapted and enhanced to estimate the system reliability in early phases, particularly to account for uncertainties associated with the unknown operation profile of the system, and unavailability of implementation artifacts and the running system.

The field of software architecture offers advanced modeling and specification techniques that describe the functional properties of software systems. These techniques are often accompanied by advanced automated tools that enable extensive analysis of the specification, but typically lack quantification and measurements. Additionally, their relationship to dependability attributes of the modeled software system is unknown.

In this talk, we present a software architecture-based approach to estimating system reliability. The approach is applicable to early stages of development, when the implementation artifacts are not yet available and the exact operation profile is unknown. The uncertainty of the operation profile is modeled using stochastic processes with unknown parameters. The approach is compositional in nature: the overall reliability of the system is estimated as a function of the reliability of its constituent components and their (complex) interactions. In turn, the reliability of individual components is estimated using standard modeling and specification mechanisms of software architectures. Additionally, sensitivity analyses enabled by our reliability model prescribe defect mitigation strategies in a cost-effective manner.

Biography

Roshanak Roshandel is a PhD candidate at the Computer Science department at the University of Southern California. She received a BS degree in Computer Science from Eastern Michigan University in 1998, and a MS in Computer Science from University of Southern California in 2002. She is a research assistant in the Software Architecture Research Group at the Center for Software Engineering under the supervision of Prof. Nenad Medvidovic. She is also a part-time software engineer with NASA's Jet Propulsion Laboratory. Her research interests are in behavioral modeling of software architectures, software dependability modeling, and design and development of distributed data intensive systems.

URL: http://sunset.usc.edu/~rroshand

Part I: A Case Study in the Participatory Design of a Collaborative Science-Based Learning Environment

Educational technology research studies have found computer and software technologies to be underutilized in U.S. classrooms. In general, many teachers have had difficulty integrating computer and software technologies into learning activities and classroom curriculums because specific technologies are ill-suited to their needs, or they lack the ability to make effective use of these technologies. In the development of commercial and business applications, participatory design approaches have been applied to facilitate the direct participation of users in system analysis and design. Among the benefits of participatory design include mutual learning between users and developers, envisionment of software products and their use contexts, empowerment of users in analysis and design, grounding of design in the practices of users, and growth of users as designers and champions of technology. In the context of educational technology development, these similar consequences of participatory design may lead to more appropriate and effective education systems as well as greater capacities by teachers to apply and integrate educational systems into their teaching and classroom practices.

Part II: Scientist-Centered Graph-Based Models of Scientific Knowledge

In recent years, semantic graphs have garnered a tremendous amount of attention among the computer science community. A bulk of this research has focused on the development of efficient data representations and their storage and operations on databases. Yet, the general notion of semantic graphs has a strong human element in that they represent and store information in forms inherent and meaningful to users. Semantic graphs may be applied within scientific computing environments as central knowledge representations that would be both useful to and usable by scientists.

At the Pacific Northwest National Laboratory, Dr. Chin is researching and developing visual models and paradigms that will allow scientists to capture and represent conceptual models in a computational form that may linked to and integrated with scientific data sets and applications. Captured conceptual models may be logical in conveying how individual concepts tie together to form a higher theory, analytical in conveying intermediate or final analysis results, or temporal in describing the experimental process in which concepts are physically and computationally explored. In this presentation, Dr. Chin describes and contrasts different research and development systems that allow scientists to capture and interact with computational graph-based models of scientific knowledge. Through these examples, Dr. Chin explores and examines ways in which researchers may graphically encode and apply scientific theory and practice on computer systems.

Capacitor mismatch is a fundamental limitation on the performance of data converters implemented using switched-capacitor circuit techniques. Recent developments in the field (mismatch-shaping algorithms) enable the circuit designer to transform traditional harmonic-distortion errors into noise-like errors of substantially the same overall power. Oversampling and noise-shaping techniques may be used to suppress the noise intensity in the signal band, but the CMOS technology's matching properties is clearly the fundamental factor that determines the overall performance. It will be shown that it is primarily the total capacitance, and not the segmentation of the matched capacitors, that determines the overall noise power. Consequently, it is found that (when some fairly lenient requirements are met) the mismatch-induced noise is reduced by 3dB every time the total capacitance of the matched capacitors is doubled. This is interesting because we know that the thermal (kT/C) noise is also reduced by 3dB when the capacitance is doubled. Hence, the ratio of these two dominating errors largely does not depend on the size of the matched capacitors, and it is to some extent beyond the circuit designer's control. The talk will quantify this important observation and address the following questions: 1) what determines the ratio of these two dominating noise sources; 2) what is the ratio today, and will it change in the future; 3) what can the circuit designer do to suppress the mismatch-induced noise. The answers may surprise you!

Biography

Jesper Steensgaard, Ph.D. in EE from the Technical University of Denmark (1999). 1997-1998: visiting scholar at Oregon State University. 2000-2001: assistant professor at Columbia University in the City of New York. 2001-2002: senior scientist with Microsemi Inc designing micro-power products. 2003-present: founder and manager of ESION LLC, and co-founder and manager of iCoustics LLC. Jesper Steensgaard specializes in the design/optimization of delta-sigma data converters, decimation/interpolation filters, class-D amplifiers, and sensor interfaces. He holds 7 US patents.

When most current applications return answers, many users do not know what information sources were used, when they were updated, how reliable the source was, or what information was looked up versus derived. Many users also do not know how answers were derived.

In this talk, we first show examples of explanations helping users to understand and trust system answers. Then we introduce the Inference Web (IW), our solution that enables explainable systems. IW aims to take opaque query answers and make the answers more transparent by providing infrastructure for presenting and managing explanations. The explanations include information concerning where answers came from (knowledge provenance) and how they were derived (or retrieved). The infrastructure includes: IWBase -- an extensible web-based registry containing details about information sources, reasoners, languages, and rewrite rules; PML -- the Proof Markup Language, an interlingua representation for justifications of results produced by software systems; and a comprehensive tool suite for browsing, checking and abstracting proofs, and explaining answers through dialogues with users.

Finally, we report on current Inference Web applications including details about two of these applications: explaining extraction as inference in support of IBM's Unstructured Information Management Architecture (UIMA) effort, and explaining task processing as inference in support of DARPA PAL's CALO personal assistant project.

Biography

Paulo Pinheiro da Silva has been a Postdoctoral Fellow of the Knowledge Systems, Artificial Intelligence Laboratory at Stanford University since 2002. During his Ph.D. study, Paulo has developed UMLi -- the Unified Modeling Language for Interactive Systems along with methods and tools aiming to facilitate systematic development of interactive systems.

Currently, Paulo is a co-technical leader of the Inference Web project, which aims to provide infrastructure for explaining answers from software systems. Inference Web is sponsored by a number of DARPA and ARDA projects. He is a former member of the Grupo de Banco de Dados of UFMG, Brazil, a leading research group on information extraction and information retrieve in Latin America. He is a former member of the Information Management Group of the University of Manchester, UK, a leading research group on Semantic Web, Grid and e-Science solutions in Europe.

Paulo received his college degree in Math and his M.Sc. degree in computer science both from the Federal University of Minas Gerais, Brazil, and his Ph.D. in computer science from Manchester University, UK.

Early, accurate detection of disease outbreaks, whether natural (e.g., West Nile virus and SARS) or bioterrorist-induced (e.g., anthrax and smallpox), is a critically important problem today. We need to detect outbreaks as early as possible in order to provide the best response and treatment, as well as improve the chances of identifying the source, whether natural or bioterroristic. An analysis of one bioagent release scenario estimated that as many as 30,000 people per day could die. The induced long-term economic costs were estimated to be as high as 250 million dollars per hour of the outbreak (Kaufmann 1997, Wagner 2001). Early detection could dramatically reduce these losses.

This need for early, accurate outbreak detection has led to the development of a new field called syndromic surveillance, which involves the collection and analysis of data which precede diagnosis in order to raise reliable alerts that merit public health investigation. This talk will give an overview of the field of syndromic surveillance and also describe the important role that data mining plays in this field. In particular, there is a need for outbreak detection algorithms that are capable of combining spatial, temporal, demographic, and symptomatic information in order to improve the timeliness and accuracy of outbreak detection. This talk will describe two such multivariate algorithms -- the What's Strange About Recent Events (WSARE) algorithm, a non-specific detection algorithm that examines case-level data for any anomalous patterns, and the Population-wide Anomaly Detection and Assessment (PANDA) algorithm, an algorithm designed to detect a large-scale, aerosolized release of an anthrax-like bioagent.

Biography

Weng-Keen Wong is a Postdoctoral Associate at the Realtime Outbreak and Disease Surveillance Laboratory at the University of Pittsburgh. He is currently involved in the Bayesian modeling for biosurveillance project. In January 2004, he completed his Ph.D. in Computer Science at Carnegie Mellon University. His dissertation topic is Data Mining for Early Detection of Disease Outbreaks, with the main focus of the thesis being the What's Strange About Recent Events (WSARE) algorithm. Originally from Vancouver, British Columbia, Weng-Keen received his B.Sc. from the University of British Columbia in 1997 and his M.Sc. from Carnegie Mellon University in 2001.

At first glance it seems absurd that busy people doing important jobs should want their computers to interrupt them. Interruptions are disruptive and people need to concentrate to make good decisions. However, successful job performance also frequently depends on people's abilities to (a) constantly monitor their dynamically changing information environments, (b) collaborate and communicate with other people in the system, and (c) supervise background autonomous services. Automated monitoring and alerting systems minimize the need to constantly monitor, but they induce alerts that may interrupt other activities. People do not perform sustained, simultaneous, multi-channel sampling well; however, they have great capacity to manage concurrent activities when given specific kinds of interface support. Empirical research with human subjects has discovered that HCI design approaches based on negotiation-based coordination support are best for enabling people to work effectively during interruptions. These results are reviewed and discussed with examples of real-world systems.

Biography

Dr. Daniel C. McFarlane is a computer scientist at Lockheed Martin Advanced Technology Laboratories (LM-ATL) in Cherry Hill, NJ. His research focus is in software architecture for human-centric artificial intelligence support systems. He is the principal investigator for the HAIL project (Human Alerting and Interruption Logistics) – a human-centric, open architecture, software component product for management of human alerting and interruption (on-going transition into the US Navy's Aegis Weapon System). Dr. McFarlane has 14.5 years of research experience at Artificial Intelligence Laboratories – 5 at LM ATL, and 9.5 at the Naval Research Laboratory in Washington, DC. He received a D.Sc. in computer science from George Washington University in 1998; an M.S. in computer science from the University of Oregon in 1990; and a B.S. in Communications from Cornell University in 1987. He is a member of ACM, Sigma Xi, and Phi Kappa Phi.

We consider wireless networks with hard energy constraints, where reducing the energy consumption becomes the most important network design consideration. An example of this type of network is a sensor network, where each node is powered by a non-rechargeable battery. Since all layers of the protocol stack affect the energy per bit consumed in its end-to-end transmission, energy minimization requires a joint design across all protocol layers as well as the underlying hardware, where the energy is actually expended.

We show that the cross-layer design across the hardware, link, MAC, and routing layers is a beneficial and feasible approach to energy efficient wireless networking. We consider both the interference-free case (with TDMA-based orthogonal MAC) and the interference-limited case (with non-orthogonal MAC). For the first case, we start with a point-to-point link, where we show that dramatic energy savings is possible when the tradeoff between the transmission energy and the circuit processing energy is explored. The results tell us that for short-range applications, bursty transmissions minimize total energy consumption. We next consider a multiple access scenario, where multiple sensor nodes are sending data to a central node. We propose a variable-length TDMA scheme to minimize the total energy consumption through a joint design of the MAC and link layers. We demonstrate the energy savings of this approach relative to a non-optimized design. We then extend our analysis to a joint design optimized across the link, MAC, and routing layers considering the hardware processing energy. We show that if link adaptation is not allowed, the energy minimization problem is a LP problem and can be efficiently solved. The solution tells us how to optimally route the traffic to minimize the total energy consumption across the network. If link adaptation is allowed, the total energy consumption can be further reduced. The optimization is no longer convex in this case, but can be relaxed to a convex problem where efficient algorithms exist to obtain a near-optimal solution. For the energy-efficient TDMA solution we obtained, we show that there exists optimal scheduling or ordering of the time slot assignments to minimize the packet delay. In the link layer, we also show that by allowing multiple nodes to cooperate, we can construct virtual MIMO systems to reduce both energy and delay. For the interference-limited case with non-orthogonal channel usage, we decompose the non-convex cross-layer problem into two sub-problems: link scheduling with heuristics, and optimal rate adaptation with routing. We’ll show how iterations between these two sub-problems lead to an energy-efficient solution. The talk will conclude with a brief discussion of joint estimation problems in sensor networks, where multiple sensors cooperate in an energy-efficient manner to estimate an unknown signal. The results will lead to a strategy where sensors with bad observation quality or bad transmission channels should be silenced to save energy.

Biography

Shuguang Cui received the B.Eng. degree in Radio Engineering with the highest distinction from Beijing University of Posts and Telecommunications, Beijing, China, in 1997, and the M.Eng. degree in Electrical Engineering from McMaster University, Hamilton, Canada, in 2000. He is currently working toward the Ph.D. degree in Electrical Engineering at Stanford University, California, USA. From 1997 to 1998 he worked at Hewlett-Packard, Beijing, P.R.China, as a system engineer. In the summer of 2003, he worked at National Semiconductor, Santa Clara, CA, as a wirless system researcher. His current research interests include cross-layer optimization for energy-constrained wireless networks, hardware and system synergies for high-performance wireless radios, and general communication theories. He is the winner of the NSERC graduate fellowship from the National Science and Engineering Research Council of Canada and the Canadian Wireless Telecommunications Association (CWTA) graduate scholarship.

The iris is a valuable biometric for use in identification, with distinctive texture that remains stable throughout life. The striations, filaments, and rings that make up the iris pattern are very unique to each person. Because of its uniqueness to an individual, it can provide identification with very high confidence through large databases. Compared with other biometric features such as face and fingerprint, iris patterns are more stable and reliable. This talk will first explain what iris recognition is. Then we introduce the One-Dimensional Iris Recognition System. We will discuss the pros and cons of this system. Moreover, we will use this system for performance analysis of partial iris recognition.

The One-Dimensional Iris Recognition System generates one-dimensional signatures to rank iris pattern similarities. It is translation, rotation, illumination and scale invariant. It is very different from traditional iris recognition systems which typically use two-dimensional iris patterns/codes/signatures for iris identification and recognition and require circular rotation for pattern matching purposes. Also, this approach uses the Du measure as a matching mechanism, and generates a set of the most probable matches instead of only the best match.

Currently, iris recognition systems require a cooperative subject, who willingly stares into a camera for a few seconds. Under these conditions, the iris image obtained the maximum amount of iris information. On the other hand, for a noncooperative subject who may be facing away from the camera, only a portion of the iris information may be captured (a partial iris). Partial iris recognition algorithms would be very important in surveillance applications. Little research has been performed in this area. In this talk, we study the performance of the use of a partial iris for recognition. A partial iris performance analysis system based on the one-dimensional approach to iris identification is introduced. Using this system, we quantitatively analyze the relationship between the accuracy rate and the portion of the iris patterns. We will discuss some interesting experimental results.

This talk is made for the general public (who have no idea about iris recognition) as well as for the professionals (in biometrics/iris recognition area).

Biography

Eliza Yingzi Du is an Assistant Research Professor and a founding member of the Center for Biometric Signal Processing at the United States Naval Academy (2003-present). She earned her Ph.D. in Electrical Engineering at the Univ. of Maryland, Baltimore County in 2003. Her research interest is in the field of digital signal/image processing and communication. She has invented a One-dimensional Iris Recognition System, created Du Measure (a hyperspetral similarity measurement method named after her), designed an automatic color thresholding scheme, developed an automatic text detection and restoration system, and designed the network protection and disaster prevention schemes for international submarine optical networks. Her research has been funded by the National Security Agency (NSA), Office of Research Lab (ONR), Naval Research Lab (NRL), and National Science Foundation (NSF). She has been serving as an active reviewer for over a dozen of journals.

Many visual perception problems, including object tracking or scene reconstruction can be formulated as inference using non-linear models, defined over high-dimensional state spaces. Suboptimal modeling, model image matching ambiguities or occlusion, often lead to representations that are weakly constrained by the image data. Therefore, robust solutions typically involve estimating and propagating highly multimodal posterior state distributions. Trapping in unrepresentative peaks represents a significant difficulty for model state inference, and it is important to locate and, for dynamic scenes, track a set of representative configurations, over time. Learning models well-adapted for the task but with good generalization power is also a necessary step, in order to reduce the search complexity, and obtain reliable estimates. In the talk, I will present inference and learning algorithms, and demonstrate these on the problem of monocular tracking and 3D motion reconstruction of a 40-dimensional articulated human model. I will focus on both generative and discriminative algorithms, discuss their assumptions and propagation rules and show how these can be used independently, but also combined in robust trackers that can efficiently self-initialize and recover from failure.

Biography

Cristian Sminchisescu received an Engineering Diploma and a MSc. from the "Politehnica" University of Bucharest, in Computer Science and Systems Simulation and Modeling, in 1996 and 1997 respectively. He has obtained a MSc. in Computer Science from Rutgers University in 1999 and a PhD. in Computer Science and Applied Mathematics (with specialization in Imaging, Vision and Robotics), from INRIA (Institute National Politehnique de Grenoble) in 2002. He has been a fellow in the Artificial Intelligence Laboratory at the University of Toronto since January 2002, he holds a visiting faculty position at Rutgers and a research faculty position at TTI-C. He has been in the program committee for ECCV and CVPR 2004 and CVPR 2005 and serves as a regular reviewer for journals and conferences including SIGGRAPH, IEEE PAMI or ICCV. His research interests are in computer vision, machine learning and computer graphics.

Wireline communication at 10Gb/s and beyond faces serious challenges. Channel impairments such as attenuation of electrical transmission line at high frequency and modal dispersion of multi-mode optical fiber degrade received signal quality significantly and result in bit error rates as large as 10-2. Adaptive equalization is a practical technique for reducing the channel response impacts and for enabling higher data rates and longer transmission distances.

In the first part of the talk, the impact of jitter on link reliability is discussed. An analytical treatment of data dependent jitter that has resulted in design insights for jitter minimization will be presented. The study provides basic theoretical understanding of previously measurement-based roots of data dependent jitter distribution. It also enables prediction of jitter impacts of general LTI systems. Experimental data verify accuracy of predictions for various components and demonstrate less than 7.5% error.

In the second part, a new approach for adaptive equalization is introduced. An eye-opening monitor architecture is presented that leverages high-speed integrated analog signal processing to evaluate the quality of the received signal. It generates a two-dimensional error map, which is directly correlated to the eye diagram shape. The evaluation result can be used as a cost function for the coefficient optimization algorithm of the equalizer. Measurements of an integrated CMOS prototype demonstrate up to 68dB error dynamic range at 10Gb/s input rate.

To achieve higher integration density and performance, silicon technology has been aggressively scaled for more than 40 years. As silicon devices are scaled down to sub-100nm regimes, power dissipation and process variations are becoming major challenges for circuit design. For continued scaling of technology, these issues must be addressed at all levels of abstraction.

The inter-die process variations, coupled with the intra-die variations can result in significant delay spread in logic circuits and even functional failures in SRAM cells. This talk will primarily focus on statistical design methodologies for yield enhancement in SRAM. A statistical modeling approach will be developed to model failure probability of an SRAM cell and the yield of an SRAM array. Based on the developed models, a methodology to statistically design the SRAM cell and the memory organization will be presented. The developed methodology can be used at an early stage of a design cycle to enhance memory yield in nano-meter regimes. The proposed design approach provides, for the first time, an integrated device, circuit, and architectural level statistical optimization strategy for SRAM. In addition to the statistical design, an adaptive body biasing technique will be presented to further improve the yield of SRAM. The last part of the talk will discuss the speaker’s research vision and future research directions.

Biography

Hamid Mahmoodi received the B.S. degree in electrical engineering from Iran University of Science and Technology, Tehran, Iran, in 1998 and the M.S. degree in electrical and computer engineering from the University of Tehran, Iran, in 2000. He is a PhD candidate in electrical and computer engineering at Purdue University, West Lafayette, IN. His research interests include low-power, robust, and high-performance circuit design for nano-scale bulk CMOS and SOI technologies. He has more than 35 refereed publications in journals and conferences. He was a recipient of the Best Paper Award of the 2004 International Conference on Computer Design.

Wireless single/multi-channel telemetric systems have always been of great interest to researchers in the biology and neurophysiology communities due to their advantage of simultaneous recording and transmission of one or more physiological parameters. Although the emergence of high-quality surface mount electrical components in the past has remarkably facilitated the implementation of such systems, the majority of the current recording devices still have either prohibitively large dimensions and weight, or high power consumption that makes them impractical for general-purpose low-power applications. Combining application-specific integrated circuit (ASIC) design techniques with micromachined biopotential recording electrode technology can significantly reduce the overall dimensions, weight, and power consumption of such systems, offering low-power multichannel radio-telemetry recording devices that can be used in closed-loop neural prosthesis to study either the peripheral or central nervous system at the cellular level. In the first part of this seminar, I will briefly talk about an ultra-light hybrid telemetric system fabricated on a foldable flexible polyimide substrate, and will demonstrate its functionality via simultaneous 2-channel wireless in vivo flight muscle EMG recordings in airborne biological hosts. The rest of the talk will then focus on the design, implementation, testing, and performance characterization of wireless FM recording microsystems-on-chip to remotely monitor input biopotentials recorded by micromachined penetrating silicon-based or nerve regeneration polyimide-substrate microprobes. The functionality of the fabricated devices in an experimental procedure will be demonstrated via single channel wireless in vivo recording of spontaneous neural activities in the auditory cortex of awake primates at different transmission ranges up to 0.5m. These devices serve as valuable test vehicles to obtain a fundamental knowledge of numerous limitations and trade-offs involved in the design of wireless integrated recording biomicrosystems.

Biography

Pedram Mohseni (S’94, M’05) was born in Tehran, Iran in 1974. He received the B.S. degree in Electrical Engineering from Sharif University of Technology, Tehran, Iran in 1996 and the M.S. degree in Electrical Engineering from the University of Michigan, Ann Arbor in 1999, where he is currently working towards the Ph.D. degree. From June-December 1998 he worked at Canopus Systems Inc., Ann Arbor, MI, developing signal processing algorithms for the multistage decimation and filtering of highly oversampled outputs of sigma-delta converters in a closed-loop microgravity microaccelerometer system. Since 2000, he has been with the Center for Wireless Integrated MicroSystems (WIMS), University of Michigan, where he is currently developing bi-directional wireless multichannel microsystems for biopotential recording applications. His research interests include analog/mixed-signal/RF integrated circuits for low-power telemetric neural interfaces. He has authored nine papers in refereed IEEE journals and conferences and has served as a technical reviewer for IEEE JOURNAL OF SOLID-STATE CIRCUITS, IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, and IEEE SENSORS JOURNAL since 2002. He is a member of the international panel of reviewers for the journal of MEDICAL SCIENCE MONITOR, the IEEE Solid-State Circuits, and the IEEE Engineering in Medicine and Biology societies.

While some substantial progress has been made in improving the way users can access interactive software systems, developing applications that effectively support users’ goals still requires considerable expertise that cannot be expected from most users. On the other hand, only a substantial involvement of users in the software development or adaptation process will ensure that users’ needs are met by a software system.

It was the objective of the European Network of Excellence EUD-Net (End-User Development Network) to clarify basic notions in end-user development and to identify and classify promising research lines in this domain. A consortium of 16 partners from 7 different European countries collaborated on this topic between 2000 and 2004. One of the main outcomes was a research agenda und roadmap for EUD, which I will summarize at the beginning of the talk.

Based on this, I will introduce our work on end-user modeling of interactive multimedia applications. This work is based on an object-oriented modeling approach by using and specializing the standard modeling language UML. Such a model-based approach is useful for end-user development because it allows people to focus on the main concepts without being confused by many low-level details.

Biography

Prof. Dr. Gregor Engels holds the chair of Information Systems at the University of Paderborn (Germany) since 1997. After his Ph.D. in Computer Science in 1986, he worked at several German universities, before he got the chair for Software Engineering at the University of Leiden (The Netherlands) between 1991 and 1997.

His research interests comprise visual, object-oriented modeling techniques and their applications in multimedia or web-based systems, software process modeling approaches as well as syntactic and semantic language engineering approaches based on graph transformation systems. He has published more than 140 scientific papers. He is regularly member of the program committee of major conferences in software engineering, visual languages and modeling approaches. Currently, he is chair of the steering committee of the VL/HCC – Visual Languages and Human-Centric Computing conference series.

More details can be found on his webpage http://www.uni-paderborn.de/cs/engels.html.

Sub-wavelength lithography processes are introducing new yield loss mechanisms at a rate, magnitude, and complexity large enough to offset “cosmetic” improvements in tools and methodologies. If EDA companies are to assist the semiconductor industry at the 90nm and 65nm nodes and below, there must be profound changes to existing tools and new flows have to be established.

In particular, there is a need for new technologies that allow designers to consider and optimize for manufacturing at each stage of the design, verification, tapeout and test process. One of the most fundamental changes is the integration of DFM and DFT flows, especially in closing the loop between manufacturing and design. DFM-oriented test uses DFM rules, recommendations, and guidelines to optimize test quality. Analysis of test data from manufacturing test creates a true goldmine of information for calibrating today’s largely qualitative DFM rules and computing yield sensitivity functions.

Biography

Janusz Rajski holds Ph.D. degree from the Poznan University of Technology, Poland. He is a Chief Scientist and Director of DFT Engineering at Mentor Graphics Corporation. He has published over 100 technical papers and holds seventeen US and international patents in the area of design for test. He is also the principal inventor of Embedded Deterministic Test (EDT™) technology and chief architect of the first commercial test compression product TestKompress™. Janusz received a number of awards including the 1993 Best Paper Award for a paper on logic synthesis published in IEEE Transactions on CAD, VTS 1995 and 1998 Best Paper Awards for papers on embedded test, and ITC 1999 and 2004 Honorable Mention Awards for papers on Logic BIST and Embedded Deterministic Test. In 2003 the President of Poland awarded the title of Professor of Sciences to Janusz for his fundamental contributions in the area of design and test of digital circuits and systems.

As the demand for bandwidth continues to increase, aggregation of signals in switches, routers and long-haul optical networks become a potential bottleneck. Serial transceivers are deployed at rates from 1-10Gbit/s to switch and route massive amounts of information throughout the data center and backbone. This approach can yield enormous cross-sectional bandwidth, but is not without its own limitations and problems.

This lecture is at a tutorial level and covers basic elements of the design of multi-gigabit-per-second serial transceivers for use in backplane and optical applications. Sources of signal impairment, such as dielectric losses, reflections at connectors, crosstalk, and package limitations will be addressed. Channel modeling and measurement techniques will also be presented. Modulation and equalization techniques to overcome the bandwidth limitations will be covered, as well as practical considerations for the integration of multiple channels onto a large digital ASIC. An overview of clock-and-data-recovery circuits will be given, and first-order circuit designs will be discussed. Finally, all aspects of the design of a serial transceiver will be reviewed in the context of a real-world example.

Biography

Aaron Buchwald was born in Ames, IA, in 1960. He received the BSEE degree from the University of Iowa in 1982, and the MS and Ph.D degrees in electrical engineering from the University of California, Los Angeles, in 1984 and 1993. He has worked for Hughes Aircraft in El Segundo, CA, and Siemens in Munich, Germany. He was also an Assistant Professor at The Hong Kong University of Science and Technology. He joined Broadcom Corporation in 1994 as the first member of the analog group. In 2003 he founded Mobius Semiconductor, an early-phase start-up in Irvine CA focusing on mixed-signal and analog technology advancements for next-generation communication physical-layer transceivers.

When multi-tasking among proactive systems, users often experience interruption overload, which causes considerable, negative impact on their performance and affective state. This burgeoning epidemic of interruption occurs not only on the desktop, but also in control rooms, aviation cockpits, and in-vehicle displays. In this talk, I will discuss our ongoing theoretical and systems development work aimed at mitigating the negative impact of interruption. Our theoretical work has produced a systematic method for predicting opportune (least disruptive) moments for interrupting users engaged in tasks. In our systems development work, we have made significant progress towards developing a theoretically-grounded system that automates the process of deferring information until opportune moments during task execution. Our work links theories of interruption from Cognitive Psychology with building interruption management systems in Computer Science.

Biography

Dr. Bailey earned a B.S. in Computer Science from Purdue University in 1993, an M.S. from the University of Minnesota in 1997, and a Ph.D. from the University of Minnesota in 2002. Dr. Bailey has been an Assistant Professor in the Department of Computer Science at the University of Illinois-Urbana since 2002. His research interests include developing computational tools that better support human creativity and storytelling, interfaces for pervasive computing, and developing computational systems that better manage human attention. Dr. Bailey’s multi-disciplinary research efforts have been formally recognized with affiliate academic appointments in Aviation Psychology, the Beckman Institute, and the Graduate School of Library and Information Science at UIUC.

UIs have proven to be a fertile ground for applying artificial intelligence techniques to build smarter, more usable interfaces.  In this talk, I present two ongoing projects that illustrate the possibilities of using machine learning in the user interface.

Smart Tasks aims to help end users automate repetitive tasks in Lotus Workplace using a programming by demonstration paradigm.  Users demonstrate how to complete one or more instances of the task, and the system learns how to automate the remainder.  Our Smart Editor demo shows how this technology applied to word processing automates repetitive style formatting tasks in documents.

Email-based Activity Management aims to mitigate email overload by automatically recognizing business processes conducted over email, and tracking one's status when participating in multiple processes simultaneously.  By recognizing common patterns in email messages, EAM is able to infer a finite-state-machine model of the process, and employ this model to categorize incoming email as steps in the process.

Biography

Tessa Lau is a Research Staff Member at IBM's T.J. Watson Research Center, where she pursues research in intelligent user interfaces: using artificial intelligence to improve human-computer interaction by building tools that adapt and learn from human use.  Before joining IBM, Tessa completed a Ph.D. in computer science at the University of Washington in 2001.  In addition to her research, Tessa's interests include gadgets, games, textiles, and encouraging women in technology.

For historical reasons, the Internet protocol stack lacks of support for efficient economic mechanisms. Consequently, Internet service providers do not have incentives to invest in new technologies and offer new services valued more by users. This leads to a stagnant industry and limits the evolution of the Internet.

In this work, we investigate issues involved in pricing Internet services, with a focus on increasing the profitability of the services. We first study pricing schemes for providing differentiated Internet services. We show that when prices are inadequate in differentiating the service qualities, the resulting system may be inefficient and even unstable. To avoid such problems, dynamic pricing schemes can be designed to ensure the stability of the system and achieve socially efficient equilibrium.

We also develop a generic model for pricing Internet services jointly offered by a network of service providers. We show that non-cooperative pricing strategy may lead to unfair distribution of revenues and discourage future upgrades to networks. We then propose a fair revenue-sharing scheme based on the concept of weighted proportional fairness. We show that this new scheme encourages collaboration and leads to higher profits for the providers.

Biography

Linhai He received his PhD in electrical engineering and computer science from the University of California at Berkeley in 2004. His research interests include economics for the Internet, applications of game theory in network protocol design, and systems design for wireless networks. Currently he is a senior member of engineering staff at Lockheed Martin's Advanced Technology Labs, working on DARPA-sponsored next-generation cognitive radio project. Prior to his PhD study, he also worked at Lucent Technologies on high-speed network design.

The continuous growth in both commercial and public network traffic with various quality-of-service (QoS) requirements is calling for better service than the Internet's best effort mechanism. One of the challenging issues is to select feasible paths that satisfy the different requirements of various applications. This problem is known as QoS routing. In general, two issues are related to QoS routing: state distribution and routing strategy. Routing strategy is used to find a feasible path that meets the QoS requirements. State distribution addresses the issue of exchanging the state information throughout the network. In this talk, I will present a novel routing architecture which consists of two parts: rate-distortion analysis based link state update and a self-adaptive QoS routing algorithm.

We first address the issue of updating link state information from the perspective of information theory. Based on the rate-distortion analysis, I will present an original scheme, which outperforms the state of the art in terms of both protocol overhead and accuracy of link state information.

QoS routing is NP-complete. Hence, tackling this problem requires heuristic. A common approach is to convert this problem into a shortest path problem and solve it with existing algorithms, e.g., Bellman-Ford and Dijkstra algorithms. However, this approach suffers from either high computational complexity or low success ratio in finding the feasible paths. Hence, I will introduce a new problem, All Hops k-shortest path (AHKP). Based on the solution to AHKP, I will present efficient self-adaptive routing algorithms, which can guarantee in finding feasible paths with fairly low average computational complexity. One of their most distinguished properties is their progressive property, which is very useful in practice: they can self-adaptively minimize their computational complexity without sacrificing the performance.

Shaped Offset QPSK (SOQPSK) is a highly bandwidth-efficient constant-envelope modulation. It is currently used in satellite and telemetry communications standards. In addition to its bandwidth-efficiency, SOQPSK is attractive because it can be demodulated with a generic symbol-by-symbol OQPSK-type detector. However, this simple and suboptimum approach does not exploit the inherent memory in SOQPSK, which is a type of continuous phase modulation (CPM).

In this work, we develop a CPM signal model that accounts for all the memory in the SOQPSK system. This model gives an optimal trellis-based detector for SOQPSK and also allows it to be viewed as a code. We apply this technique to serially concatenated coding schemes with iterative detection, where SOQPSK itself constitutes the inner code. A number of reduced-complexity designs are presented. In particular, it is shown that all versions of SOQPSK, from the simplest to the most complex, can be treated with a simple, common architecture.

Biography

Erik Perrins is a graduating PhD student from the Department of Electrical and Computer Engineering at Brigham Young University. He also received his BS in 1997 and his MS in 1998 from BYU, both in electrical engineering. From 1998-2004 he was with the Advanced Technology Group of Motorola in Schaumburg, IL. Since 2004 he has been an industry consultant while concurrently finishing his studies at BYU. His research interests are digital transmission theory, modulation and coding, and signal processing.

Graphics cards for personal computers have recently undergone a radical transformation from fixed-function graphics pipelines to multi-processor, programmable architectures. Multi-processor architectures are clearly advantageous for graphics for the simple reason that graphics computations are naturally concurrent, mapping well to stateless stream processing. They therefore parallelize easily and need no random access to memory with its problematic latencies.

This paper presents Vertigo, a purely functional, Haskell-embedded language for 3D graphics and an optimizing compiler that generates graphics processor code. The language integrates procedural surface modeling, shading, and texture generation, and the compiler exploits the unusual processor architecture. The shading sub-language is based on a simple and precise semantic model, in contrast to previous shading languages. Geometry and textures are also defined via a very simple denotational semantics. The formal semantics yields not only programs that are easy to understand and reason about, but also very efficient implementation, thanks to a compiler based on partial evaluation and symbolic optimization, much in the style of Pan.

Haskell's overloading facility is extremely useful throughout Vertigo. For instance, math operators are used not just for floating point numbers, but also expressions (for differentiation and compilation), tuples, and functions. Typically, these overloadings cascade, as in the case of surfaces, which may be combined via math operators, though they are really functions over tuples of expressions on floating point numbers. Shaders may be composed with the same notational convenience. Functional dependencies are exploited for vector spaces, cross products, and derivatives.

Biography

Conal Elliott's research aims mainly at simplifying the creation of interactive synthetic media content, while preserving or improving on the flexibility and performance of mainstream programming approaches. Continuous and infinite time and space play key roles in simplifying the conceptual model and improving composability.

Conal has a PhD from Carnegie Mellon University (1990, "Extensions and Applications of Higher-Order Unification") and a BA in math from the College of Creative Studies at UC Santa Barbara. He worked at Sun Microsystems for five years and was a graphics researcher at Microsoft Research for eight years. Starting in October 2002, he spent a two-year personal sabbatical, exploring mainly compassionate (nonviolent) communication, spirit, psyche, jazz singing, and dance. See http://conal.net for more info.

Multiprocessor system-on-chip (MP-SoC) platforms are emerging as an important trend for System on Chip (SoC) design. The state of the art has reached a point where commercial designs are readily integrating in the range of 10-100 embedded functional/storage blocks in a single SoC. As a result of this enormous degree of integration, several industrial and academic research groups are striving to develop efficient communication architectures, in some cases specifically optimized for specific applications. One of the major problems associated with future SoC designs arises from non-scalable global wire delays, which typically increase exponentially or, at best, linearly by inserting repeaters. Even after repeater insertion, the delay may exceed the limit of one clock cycle (often, multiple clock cycles). According to ITRS (2003 update), “Global synchronization becomes prohibitively costly due to process variability and power dissipation, and cross-chip signaling can no longer be achieved in a single clock cycle”. Thus, system design must incorporate networking and distributed computation paradigms with communication structures designed first and then functional blocks integrated into the communication backbone.

The emerging Network on chip (NoC) design methodology is a step towards this. The practical implementation and adoption of the NoC design paradigm is faced with various unsolved issues related to design methodologies, test strategies, and dedicated CAD tools. The focus of this seminar is on design aspects and architectural issues of this new paradigm. The first part presents a quantitative comparison of various candidate architectures in regards to data rates, latency, silicon area overhead, and energy dissipation. This comparison is based on a set of metrics that establish a useful basis for the optimal evaluation and selection of interconnect infrastructures for large and complex SoCs. The second part focuses on the timing characteristics of NoC architectures. Through detailed circuit design and timing analysis, this research has established that different NoC architectures proposed to date are guaranteed to achieve the high-performance clock cycle requirements in a given CMOS technology, usually specified in normalized units of FO4 (Fan out of 4) delays. The presentation will conclude with future research directions, namely incorporation of on-chip error control coding mechanisms, fault tolerant architectures and test methodologies supporting the NoC paradigm.

Biography

Partha Pande is finishing his PhD in VLSI design at the department of Electrical and Computer Engineering, University of British Columbia. His PhD thesis has evolved around the topic of “Networks on Chip”. Previously Partha obtained his M.Sc in Computer Science from the National University of Singapore in 2001 and Bachelor degree in Electronics and Communication Engineering from the Calcutta University, India in 1997. After his Bachelor degree he worked in Industry for couple of years as Digital System Design Engineer. At the end of 1999 Partha came back to academia to pursue higher studies.

Partha has won several national scholarships from the Govt. of India for academic excellence. In addition to this he received the International Graduate student scholarship from the National University of Singapore. He has got several publications in refereed journals and conferences.

Community computing presents a number of challenges for designers of information technology. The end user population is diverse with respect to background and motivation as well as time and other resources. Users’ learning about the technology and applications of community computing is often informal and ad hoc, occasioned by the need to accomplish a real world goal. In this talk I will describe the requirements for informal learning in community computing and argue that the minimalist design model—originally developed in the context of personal computing—can be usefully generalized to address these needs. I will illustrate the constructs of minimalism and how they can be applied to community computing through a number of exploratory research projects conducted over the past decade. I will conclude with a brief discussion of the more general implications for minimalism as an orienting framework for collaborative computing.

Biography

Mary Beth Rosson is Professor of Information Sciences and Technology at Pennsylvania State University. She received a PhD in experimental psychology in 1982 from the University of Texas. Prior to joining the new School of IST at Penn State in 2003, she was a professor of computer science at Virginia Tech for 10 years and a research staff member and manager at IBM’s T. J. Watson Research Center for 11 years. Rosson’s research interests include scenario-based design and evaluation, the use of network technology to support collaboration, especially in learning contexts, and the psychological issues associated with use of high-level programming languages and tools. She is author of Usability Engineering: Scenario-Based Development of Human-Computer Interaction (Morgan Kaufmann, 2002), Instructor’s Guide to Object-Oriented Analysis and Design with Application (Benjamin Cummings, 1994), as well as numerous articles, book chapters, and tutorials. More information is available at http://ist.psu.edu/rosson.

Quantum computation is a recent approach to information processing which is based on quantum mechanics rather than classical physics. Information is represented by quantum bits (qubits) which correspond to quantum states such as photon polarization. Because of the superposition property of these states, n qubits can store up to 2n binary words simultaneously, suggesting a type of massive parallelism. Quantum states also allow powerful forms of interaction such as interference and entanglement, which have no counterparts in classical (non-quantum) computation. Quantum computers can solve a few important and hitherto intractable problems such as prime factorization of large numbers. Novel forms of highly secure communication are also possible. In practice, however, quantum computing devices and circuits are extremely difficult to construct, since they are nanoscale in size and operate at extremely low energy levels. Consequently, they have many more failure modes than classical circuits For example, quantum signal states are inherently unstable and tend to decay rapidly due to interaction with the environment (decoherence). Quantum gate operations are defined by continuous parameters that allow small errors to arise and propagate to other gates. Furthermore, state measurement is probabilistic and the measurement process itself affects the state being measured. This talk will review the history and development of quantum circuits, with emphasis on their relation to classical computer circuits and their fault-tolerance requirements.

Biography

John P. Hayes is the Claude E. Shannon Professor of Engineering Science at the University of Michigan, where he teaches and conducts research in the areas of VLSI CAD, fault tolerance, mobile embedded systems, and quantum computing. He received the B.E. degree from the National University of Ireland, and the M.S. and Ph.D. degrees from the University of Illinois. At Illinois he participated in the design of the ILLIAC III computer. He spent ten years with the University of Southern California, before moving to the University of Michigan. Hayes was the founding director of Michigan's Advanced Computer Architecture Laboratory (ACAL). He is the author of numerous technical papers, several patents, and five books, including Computer Architecture and Organization, (3rd ed., McGraw-Hill, 1998). Hayes is a Fellow of IEEE and ACM. He received the University of Michigan’s Distinguished Faculty Achievement Award in 1999 and the Humboldt Foundation’s Research Award in 2004.

Abundance of cheap embedded sensors equipped with processing, communication & actuation capabilities has created considerable interest in sensor network applications. In order to solve the fundamental problems underlying these systems, multi-disciplinary approaches are required that can integrate between different fields such as sensing, communication, estimation and control. In this talk we develop new theoretical foundations for mobile sensor networks running real-time applications with the aim of integrating communication and estimation/control. On the communication side, we will show that the communication protocols and designs suitable for other already-existing applications like data networks may not be entirely applicable for estimation and control of a rapidly changing dynamical system. Then we develop new design theories to improve the performance and stability of these systems. For instance, we examine the role of a cross-layer feedback and will see how the optimum design should provide a balance between information loss and communication noise in the absence of such a feedback. On the control side, we will consider decentralized control of sensor trajectories. There we provide new design strategies that would consider the trade-offs between sensing and communication.

Biography

Yasamin Mostofi is a postdoctoral scholar in the department of Electrical Engineering at California Institute of Technology. She received her Bachelor in Electrical Engineering from Sharif University of Technology in 1997. She then received her MS and PhD in Electrical Engineering from Stanford University in 1999 and 2003. Her research areas of interest include sensor networks, communications and control and dynamical systems.

In this talk I describe a new concept to deal with persistent information in a functional logic language. First, I give a short survey on the declarative multi-paradigm language Curry that integrates the main features of functional, logic, and concurrent programming. Then, I introduce the concept of dynamic predicates for dealing with persistent information in functional logic programs. The definition of a dynamic predicate can change over time, i.e., one can add or remove facts that define this predicate. Dynamic predicates are easy to use and have a clear semantics that does not depend on the particular (demand-driven) evaluation strategy of the underlying implementation. In particular, the concept is not based on (unsafe) side effects so that the order of evaluation does not influence the computed results. The latter is an essential requirement in non-strict languages.

Dynamic predicates can also be persistent so that their definitions are saved across invocations of programs. Thus, dynamic predicates are a lightweight alternative to the explicit use of external database systems. Moreover, they extend one of the classical application areas of logic programming to functional logic programs. Finally, I discuss the implementation of dynamic predicates.

Biography

Michael Hanus is Professor of Computer Science at the University of Kiel, Germany. His research is mainly concerned with the integration of functional and logic programming languages, the design and implementation of declarative programming languages, type systems for logic programming, analysis techniques for declarative programs, programming environments and applications of declarative languages. He has published more than eighty papers on these topics in international conference proceedings, journals and books. Currently, he is involved in the design, implementation, and application of the multi-paradigm declarative language Curry.

In this talk, I will present a novel photographic technique called dual photography, which exploits Helmholtz reciprocity to interchange the lights and cameras in a scene. With a video projector providing structured illumination, reciprocity permits us to generate pictures from the viewpoint of the projector, even though no camera was present at that location. The technique is completely image-based, requiring no knowledge of scene geometry or surface properties, and by its nature automatically includes all transport paths, including shadows, inter-reflections and caustics.

In its simplest form, the technique can be used to take photographs without a camera. I will show results of images we captured using only a projector and a photo-resistor. If the photo-resistor is replaced by a camera, we can produce a 4D dataset that allows for relighting with 2D incident illumination. Using an array of cameras we can produce a 6D slice of the 8D reflectance field that allows for relighting with arbitrary light fields. Since an array of cameras can operate in parallel without interference, whereas an array of light sources cannot, dual photography is fundamentally a more efficient way to capture such a 6D dataset than a system based on multiple projectors and one camera. As an example, I will describe how dual photography can used to capture and relight scenes.

Biography

Pradeep Sen is a PhD candidate in Electrical Engineering in the Graphics Laboratory at Stanford University. He received his BS degree in Computer and Electrical Engineering from Purdue University in 1996 and his M.S. in Electrical Engineering in 1998 in the area of electron-beam lithography. For his thesis, he developed a structure called a silhouette map and applied it to various problems in computer graphics such as texture magnification and shadow mapping. Other publications in graphics include work on real-time shading and on dual photography as presented in this talk. His interests include real-time graphics and graphics hardware, global illumination algorithms, computational photography and display technology.

Understanding noise in submicron MOS devices is an ongoing challenge in the area of mixed-signal modeling. Experimental observations show that the classical long-channel MOSFET noise formulation underestimates the drain current noise of short-channel devices by a factor often referred to as the excess noise factor. In order to predict the effects of this excess noise on amplitude and phase noise in future circuits, it is crucial to have a reliable MOSFET noise model and an accurate phase noise formulation. This presentation outlines the latest theoretical and experimental results of Stanford’s TCAD and Microwave Integrated Circuits (SMIrC) labs on this issue. Directions for further research will also be discussed.

Biography

Reza Navid received his B.S. degree in Electrical engineering in 1996 from the University of Tehran, Tehran, Iran. He received his M.S. degree also in Electrical engineering from Sharif University of Technology in 1998 with an emphasis on EM. From 1998 till 2000 he was with ParsElectric MFG Corp., Tehran, Iran, where he was working on TV tuner characterization and improvement. He entered the University of Michigan, Ann Arbor, Michigan in January 2000 as a graduate student researcher working on "MEMS for Wireless Communication". During summer 2001 he was with Maxim Integrated Products, Hillsboro, Oregon. He is now with Stanford University to pursue his Ph.D. in Electrical Engineering. His current research interests cover a wide range of questions regarding analog aspects of electronic systems, circuits and devices. He is currently working on various areas related to RF noise (both amplitude and phase noise).

During the last four years, we have been working on our WebGD framework and WebGD application generator (WebGD-GEN). These software tools allow major parts of advanced Web-based GIS/ database applications, including the map interfaces, to be created automatically. The applications generated supports the following unique features. 1) Geographical features and the data associated with them can be entered, searched, updated, and deleted from standard Web browsers. 2) The map projection used to display a particular region in the world can be automatically switched to avoid map distortion. 3) An application can be customized with configuration files, and Web scripts can be automatically generated from a relational database schema. Thus, a WbGD application can be created without any programming, except for the scripts that are hooked to the automatically generated scripts. Application specific processing can be handled by the scripts hooked. In my presentaion, the organization of a typical WebGD application and the design of the WebGD framework and WebGD-GEN are explained. I then will demonstrate several WebGD applications developed with our tools. You can find a copy of a paper describing the WebGD approach at http://web.engr.oregonstate.edu/~minoura/papers_pub/ssdbm05_9.pdf

Biography

Measured impedances of mm-wave components in the B.S. project. Designed and implemented a digital control system for a rate- gyroscope in the M.S. project. Enagaged in design and implementation of computer-control systems for fossil-fuel and nuclear power plants for six years. Obtained a Ph. D. degree by working on concurrency and recovery schemes for distributed database systems. Created many sample programs as active object systems. Currently working on Web- based applications that use databases and GIS interfaces. A licensed first-class radio engineer (Japan) and a licensed fisherman (Oregon).

In an abstract sense, CPUs have changed very little over the past couple of decades -- in many cases, 20-year-old binaries will run on new machines. However, the performance characteristics of computers systems has changed dramatically over this period of time, particularly for SMP operating-system kernels such as Linux. This talk will summarize some of these changes, and will give an overview of how a recent synchronization mechanism, namely RCU ("read-copy update") has greatly improved performance of algorithms involving read-mostly data structures in the Linux kernel.

Biography

Paul E. McKenney is a Distinguished Engineer at the Linux Technology Center at IBM Beaverton. He obtained his Ph.D. in Computer Science and Engineering from the Oregon Health and Sciences Univerity. His research interests are in the intersection of SMP performance and scalability and realtime response. He has authored more than 20 research papers in SMP algorithms, communications congestion control algorithms, realtime response, and synchronization primitives. He is a member of the IBM Academy of Technology, and has been invited to the Linux Kernel Summit three times (2003-2005). He is a co-inventor of RCU, but given that something resembling RCU has been invented independently at least five times, this might not count for all that much. He holds 19 patents. His hobbies include running and the usual house-wife-and-kids habit.

Despite stereotypes about solitary nerds in cubicles, corporate software development is a very social human activity. In recognition of this, over the past year, we've begun a new effort at Microsoft Research to combine ideas and research methods from HCI and CSCW with those from software engineering to produce new development tools. This talk provides a survey of our work so far: Team Tracks allows a new team member to quickly learn the team's source code by mining data about the team's browsing and editing habits; the Bridge makes it easy to find task-relevant context information, drawn from source history, bug databases, email, web sites, and online documents; Software Terrain Maps are designed to allow the developer to navigate around code using spatial memory rather than a web of memorized names and relationships. In short, we're studying software as if it were created by PEOPLE working TOGETHER.

Biography

Rob DeLine and Gina Venolia call their group Human Interactions In Programming to ensure that all their research is HIP. Rob DeLine's research covers several academic disciplines: HCI (Alice), software architecture (UniCon, flexible packaging) and program verification (Vault, Fugue, Spec#). Gina Venolia has been a user interface architect with Microsoft Research since 1998. She has a checkered past that includes 15 years at Apple Computer, where, among other things, she shipped one of the original applications for the Lisa computer.

Many areas within Comptuer Graphics make use of objects that have a rigid or deformable shape. Such areas include rendering, geometric modeling, and computer animation. Shape analysis allows us to better understand the geometric structures in an object, which often leads to more efficient and robust algorithms. In this talk, I will demonstrate the importance of shape analysis and understanding with three graphics applications: mesh simplification, surface parameterization, and pen-and-ink illustration of shapes.

Biography

Dr. Zhang is currently an Assistant Professor in the School of Electrical Engineering and Computer Science of Oregon State University. He obtained his Ph.D. in Computer Science from Georgia Institute of Technolgoy in 2004. His areas of research interests include: computer graphics, scientific visualization, geometric modeling, computational topology, and non-photorealistic rendering.

Despite the acclaimed enhanced time-frequency feature detailing capability of quadratic time-frequency analysis techniques in the technical literature (for example, Wigner distributions), the performance with actual signals is often quite poor due to the inability to suppress the croos-term artifacts inherent in quadratic TFAs. This seminar will present a modification of classic short-time Fourier transform (STFT) to provide the detailing capability of quadratic TFAs without the cross terms that hide extremely weak signal features. The method is based on some high-resolution 2-D linear predictive based signal modeling techniques. Examples with communication and radar will be presented.

Biography

Larry Marple received the B.S. and M.E.E. degrees from Rice University (Houston, TX) and the Engr.D. from Stanford University, all in electrical engineering. He is currently a Professor of ECE at Oregon State University, with research interests in sensor signal processing. Prior to joining OSU, he was for 11 years the Chief Scientist of the Orincon Corporation in San Diego, CA. He has 31 years of experience in the development and implementation of digital signal processing algorithms, software, and hardware for temporal, frequency, and spatial domain signals in sonar/underwater acoustics, radar, communications, intelligence, acoustic well logging, and ultrasound imaging. He is author of the text DIGITAL SPECTRAL ANALYSIS (Prentice Hall, 1987). He was elected a Fellow of the IEEE in 1989 for "contributions to the theory and application of spectral analysis in digital signal processing."

A twenty year history of the free/open software movement, from my perspective of how to contribute (in many ways!) and how to make money as well.

Biography

Randal L. Schwartz is a renowned expert on the Perl programming language (the lifeblood of the Internet), having contributed to a dozen top-selling books on the subject, and over 200 magazine articles. Schwartz runs a Perl training and consulting company (Stonehenge Consulting Services, Inc of Portland, Oregon), and is a highly sought-after speaker for his masterful stage combination of technical skill, comedic timing, and crowd rapport. Schwartz is also infamous amongst the System Administration community for his arguable 1995 criminal conviction while performing activities for the Intel Corporation, and publicly advocates for appropriate computer crime laws. His presentation about the landmark case (titled "Just Another Convicted Perl Hacker") has inspired action for computer professionals and lobbyists at computer conferences all over the world. And he's a pretty good Karaoke singer, winning contests regularly.

Most workflow management systems take a process-oriented approach to describing workflows, where the focus is on applications and how they are interconnected. This talk presents an alternative, data-centric approach, where the emphasis is on work products. A workflow description is a set of rules, where each rule describes the data produced by a workflow step, the data the step depends on, and the processes used to produce the data. A prototype workflow enactor named PIP uses this rule-based model to manage several bioinformatics projects, including the construction of a tRNA datamart and projects related to the evolution of duplicate genes. Advantages of the rule-based approach are modularity and reusability: steps in a PIP workflow can be described as objects, and new steps can inherit many of the attributes of previously developed steps.

Biography

John Conery is a Professor of Computer and Information Science at the University of Oregon, where he has been a faculty member since earning his PhD at UC Irvine in 1983. His research interests include parallel processing, computational science, and bioinformatics.

This colloquium will be a panel discussion that brings together a variety of perspectives on how to interview for an academic job. The topics covered will be: (1) What are the faculty looking for and how are they deciding which candidates have it? (2) What are some specific interviewing skills a candidate should know? (3) Tips from the trenches from recent academic job candidates. (4) Open discussion/q&a (other faculty and soon-to-be job candidates in the audience should join the fray here).

Biography

Margaret Burnett is a Professor in the School of EECS, and has been on the School's hiring committee most years since 1997. She will explain the hiring process here at OSU, with a strong emphasis on what the hiring committee (and faculty) have been looking for the past few years. Martha Chamberlin is Director of Project Development and Training, OSU Extended Campus. In the course of interviewing for her present position, she took a course on interviewing, and will share the many insights she gained from applying the course to her experience, including things like: your values, your inventory of skills, elevator speech (me in 30 seconds), preparing to sound prepared, and more. Carlos Jensen is an Assistant Professor who was on the job market last year (2004/5). He brings ideas and perspectives from the perspective of an interviewing PhD student from Georgia Tech. Alan Fern is an Assistant Professor who was on the job market in 2003/4. He brings ideas and perspectives from the perspective of an interviewing PhD student from Purdue. Weng-Keen Wong is an Assistant Professor who was on the job market last year (2004/5). He brings ideas and perspectives from the perspective of an interviewing PhD student from Carnegie Mellon. Thinh Nguyen is an Assistant Professor who was on the job market in 2003/4. He brings ideas and perspectives from the perspective of an interviewing PhD student from Berkeley.

Although gender differences in a technological world are receiving significant research attention, much of the research and practice has aimed at how society and education can impact the successes and retention of female computer science professionals--but the possibility of gender issues within problem-solving software has received no attention. To begin addressing this issue we conducted an extensive literature search outlining possible gender differences which may effect end users engaged in computer-based problem-solving tasks. In this talk I will focus on how gender differences in computer self-efficacy (a form of confidence) impact users' problem-solving effectiveness. Based on findings from a study looking at self-efficacy we made design changes to accommodate users with low self efficacy, and then conducted a follow-up study investigating those changes. Studying these gender HCI issues is important because without knowledge of how gender differences interact with software designers could unintentionally be making design choices which leave a large portion of the population at a disadvantage for computer-based problem-solving.

Biography

Laura Beckwith is a Ph.D. candidate at Oregon State University working with Margaret Burnett. She received her undergraduate degree in 1998 at Hobart and William Smith Colleges in central New York state. After working as a software engineer for several years in Rochester, NY she moved to Corvallis and started her masters program in computer science. Since completing her masters degree in 2002 she has been working on her PhD entitled "Gender HCI." Her primary research interests are in HCI (human-computer interaction) and end-user software engineering.

Continually monitoring the state of a dynamic system is an important problem for artificial intelligence. Dynamic Bayesian networks (DBNs) provide for compact representation of probabilistic dynamic models. However the monitoring task is extremely difficult even for well-factored DBNs. Therefore approximate monitoring algorithms are needed. One family of approximate monitoring algorithms is based on the idea of factoring the joint distribution over the state of the system into a product of distributions over factors consisting of subsets of variables. Factoring relies on the notion of weak interaction between subsystems. We identify a new notion of weak interaction called separability, and show that it leads to the property that, in order to compute the factor distributions at one point in time, only the factored distributions at the previous time point are needed. We also define an approximate form of separability. We show that separability and approximate separability lead to very good approximations for the monitoring task. Unfortunately, sometimes the factoring approach is computationally infeasible. An alternative approach to approximate monitoring is particle filtering (PF), in which the joint distribution over the state of the system is approximated by a set of samples, or particles. In high dimensional spaces, the variance of PF is high and too many particles are required to provide good performance. We improve the performance of PF by introducing factoring, maintaining particles over factors instead of the global state space. This has the effect of reducing the variance of PF and so reducing its error. Maintaining factored particles also allows us to improve PF by looking ahead to future evidence before deciding which particles to propagate, thus leading to much better accuracy.

Biography

Avi Pfeffer is Associate Professor at Computer Science at Harvard University. His research is directed towards achieving rational behavior in intelligent systems, based on the principles of probability theory, decision theory, Bayesian learning and game theory. He received his PhD in 2000 from Stanford University, where his dissertation on probabilistic reasoning received the Arthur Samuel Thesis Award. Dr Pfeffer has published technical papers on probabilistic reasoning, strategic reasoning, agent modeling, temporal reasoning, and database systems. He was awarded the NSF Career Award in 2001 for work on strategic reasoning, and the Alfred P. Sloan Foundation Research Fellowship in 2002. Dr Pfeffer serves on the editorial board of the Journal of Artificial Intelligence Research, and on the program committees of a number of leading conferences in artificial intelligence.

The talk will give an overview of environmental applications of machine learning. Example applications include earthquake prediction and modelling gene- flow between GM and conventional crops. The talk will present in more detail two general classess of ecological modelling problems, namely modelling population dynamics and modelling habitat suitability, that are often addressed by using machine learning methods. Example applications from each class will be given, such as modelling algal growth in the lagoon of Venice and modelling the habitat of brown bears in Slovenia.

Biography

Saso Dzeroski is a senior scientific associate at (and deputy head of) the Department of Knowledge Technologies, Jozef Stefan Institute, Ljubljana, Slovenia. He is also an associate professor at the Jozef Stefan International Postgraduate School. He has conducted research on a wide variety of topics within machine learning (including computational scientific discovery, computational learning theory, relational learning, inductive logic programming, reinforcement learning) and its applications (environmental sciences, life sciences and natural language processing). He has also organized a range of scientific and educational events (conferences, workshops, seminars) on the above topics and co-authored/co-edited three books and five published proceedings . He was program co-chair of ICML-99, invited speaker at ICML-02 and general chair of ICML-05 (The Twentysecond International Conference on Machine Learning).

An important challenge for machine learning is to find ways of learning quickly from very small amounts of training data. This is important for applications such as modeling the transmission of new diseases, understanding the propagation of new computer worms, modeling preferences and risk behavior in negotiations, and so on. The only way to learn from small data samples is to constrain the learning process by exploiting background knowledge. The key is to identify kinds of background knowledge that are (a) easy for experts to specify and (b) easy for algorithms to exploit. One example of such background knowledge is the causal relationships in a domain, which can be encoded in the graph structure of a Bayesian network. While causal relationships help constrain Bayesian network learning, they do not always provide enough constraint. The number of parameters that must be learned grows exponentially with the number of parents, and this means that very large amounts of data are needed. We have identified another kind of background knowledge, qualitative monotonicities, that is easy to specify and that can help address this problem. A qualitative monotonicity statement says that one variable increases (or decreases) monotonically as a function of another variable, e.g., "warmer temperatures increase the size of the mosquito population". Our research focuses on algorithms for learning from small samples constrained by qualitative monotonicity knowledge. This talk will show how qualitative monotonicity statements can be implemented using logistic regression to learn the conditional probability distributions of a Bayesian network. In particular, we present a logistic regression model in which the number of parameters and the number of constraints needed to learn these conditional probability distributions do not grow exponentially with the number of parents. Experimental results using our approach suggest faster learning from very small training sets.

Biography

Angelo Restificar is a Research Associate in the School of Electrical Engineering and Computer Science at Oregon State University in Corvallis, Oregon. He received his Ph.D. in August 2004 in Computer Science from the University of Wisconsin-Milwaukee. During his graduate work, Dr. Restificar developed a framework for learning the preferences and risk behavior of agents engaged in negotiation. He is actively involved in seeking solutions to research issues that are critical to learning systems in agent-agent and agent-human interaction – an area of research in which data sparseness could pose a serious problem for modelers. Dr. Restificar’s general research interests involve issues that span the areas of machine learning, decision theory, human-computer interaction, game theory, and natural language processing.

This talk will review recent work in our group, on circuits that combine domains usually kept separate. The first systems to be discussed are digital signal processors in which the binary waveforms used are functions of continuous time. No sampling is used in converting the signal from analog to digital form, and thus there is no aliasing of signal or of quantization error. This can result in much smaller in-band quantization error than is possible with sampling and discrete-time digital signal processing. Also to be discussed are input-output linear analog filters which are internally nonlinear, and processors in which digital waveforms are processed directly with analog circuits. The talk will conclude with a description of analog VLSI computers, which make approximate computation faster, and which can co-operate with digital computers to speed up accurate computation.

Biography

Yannis P. Tsividis received his BEE degree from the University of Minnesota, Minneapolis, and his MS and Ph.D. degrees from the University of California, Berkeley. He is Charles Batchelor Professor of Electrical Engineering at Columbia University in New York. Starting with the first fully-integrated MOS operational amplifier, which he demonstrated in 1976, he has done extensive work in analog and mixed-signal integrated circuits at the device, circuit, system, and computer simulation level. He is the recipient of the 1984 IEEE W.R.G Baker Award for the best IEEE publication, the 1986 European Solid-State Circuits Conference Best Paper Award, and the 1998 IEEE Circuits and Systems Society Guillemin-Cauer Best Paper Award. He is co-recipient of the 1987 IEEE Circuits and Systems Society Darlington Best Paper Award and the 2003 IEEE International Solid-State Circuits Conference L. Winner Outstanding Paper Award. He is a Fellow of the IEEE. Among his teaching awards is Columbia’s 2003 Presidential Award for Outstanding Teaching, and the 2005 IEEE Undergraduate Teaching Award.

The design of a 155Mb/s - 4.25Gb/s laser driver in SiGe BiCMOS is described. A large output voltage compliance range that allows DC coupling to the laser diode is achieved with a translinear pseudo-differential output driver. Active back-termination is provided at the modulation output pins. Careful design of the level shift stage affords low deterministic jitter over a very wide range of bit rates. The dynamic performance is preserved over a wide range of modulation current with a segmented driver slice scheme.

Biography

John W. Fattaruso grew up in Berkeley, CA, and received the B.S.(highest honors), M.S., and Ph.D. degrees in electrical engineering from the University of California, Berkeley, through 1986. He has been a Hertz Foundation Fellow, Teaching Associate, Research Assistant and Instructor at the University of California, Berkeley. In 1979 he worked in the Digital Signal Processing R&D group at Hewlett-Packard, Santa Clara, CA, and in 1985 he served as a consultant to Seeq Technology, San Jose, CA. Since 1987 he has been with various research and product development departments of Texas Instruments, Dallas, TX, working in many areas of analog VLSI technology. He was elected Distinguished Member of Technical Staff in 2001. His technical interests include analog and RF circuit design, circuit simulation and optimization, and numerical analysis. He currently holds 29 patents in circuit design, has authored or co-authored 20 journal papers and conference presentations, and has served on the analog program subcommittee of the ISSCC and as guest editor of the JSSC. Dr. Fattaruso is a member of Eta Kappa Nu, Tau Beta Pi and Phi Beta Kappa.

New developments in high speed CMOS serial links have been crucial in matching off-chip system bandwidth with the ever increasing on-chip demand. As transistor scaling pushes data rates higher, improving the timing precision in transceiver architectures becomes a key factor in the performance of future high speed serial links. Conventional serial links architectures typically use multi-phase clocking structures to achieve a pin bandwidth faster than the on-die logic switching speed. However, multi-phase clocking architectures suffer from significant sources of timing uncertainty, namely power supply induced jitter and static phase offset resulting from process mismatch. These jitter sources are expected to worsen in future CMOS technologies, reducing the effectiveness of such multi-phase transceiver architectures for very high data rates. In this talk, I will present a new serial link architecture which addresses the timing uncertainties caused by power supply noise and process mismatch. In contrast with multi-phase architectures, the complementary phases of an integrated LC-VCO directly drive the final output multiplexer in the transmitter, resonating the load and eliminating clock buffers, thereby reducing power dissipation, power supply induced jitter, and static phase offset. In the receiver, a similar technique is applied, where the front-end input sampler is directly driven by a different LC-VCO. The resulting power supply susceptibility is reduced by 10x and process mismatch phase error by 5x. Two test chips have been designed and fabricated in 0.13um CMOS technology, exhibiting a 20Gb/s data rate with low power/area. The architecture, circuit design, measured results, and future work will be presented.

Biography

Patrick Chiang received the B.S. degree in electrical engineering and computer sciences from the University of California, Berkeley, in 1997, and the M.S. degree in electrical engineering from Stanford University in 2001. He is currently working toward the PhD. degree in the Computer Systems Laboratory at Stanford University. In 1998, he was with Datapath Systems (now LSI Logic), working on analog front-ends for DSL chipsets. In 2000, while collaborating with UCSD, he implemented one of the first monolithic implementations of a complete chaotic transceiver on a CMOS die. In 2001, he worked with Dr. Ed Lee on two prototypes of 0.25um, 4Gb/s low power/area CMOS serial links. Recently, his doctoral work has been on the design of precision clock synthesis architectures for high speed serial links, resulting in the implementation of two 0.13um, 20Gb/s low power/area CMOS serial links. In 2004, he consulted at Telegent Systems, Sunnyvale, CA, working on various mixed-signal RF blocks. His interests are in ultra-wideband RF architectures, high speed serial links, high speed A/D conversion, circuit architectures for process variation, and MEMS / circuit interfaces.

In order for the aggressive scaling of transistors to truly benefit the performance of large and complex digital systems, the communication bandwidth between ICs must scale accordingly. However, interconnect technology does not scale as aggressively, making communication between chips the major bottleneck in overall system performance. In addition, supply voltage scaling, increasing device leakage, and increased noise make existing signaling circuits inefficient and difficult to scale. This talk will present both analog and digital enhancement techniques to mitigate scaling related issues and improve the performance of building blocks used in high-speed signaling systems. A hybrid analog/digital clock and data recovery (CDR) architecture that improves the tracking range of traditional CDRs by an order of magnitude will be presented. This CDR also employs an improved analog phase-locked loop architecture to circumvent fundamental scaling problems such as low voltages and reduced output impedances. An all-digital CDR architecture that obviates the need for any analog components while achieving error-free operation will be discussed. Several digital signal processing techniques used to achieve this performance will be presented. Finally, a digital-to-phase converter(DPC) with a resolution that exceeds the state-of-the-art DPC resolution by an order of magnitude will be reported. The proposed DPC achieves better than 100 femto-second resolution. Measurement results obtained from prototype chips that validate the proposed design techniques will be summarized.

Biography

Pavan Kumar Hanumolu received the B.E. (Hons.) degree in electrical and electronics engineering and the M.Sc. (Hons.) degree in Mathematics from the Birla Institute of Technology and Science, Pilani, India, in 1998, and the M.S. degree in electrical and computer engineering from the Worcester Polytechnic Institute, Worcester, MA, in 2001. He is currently working toward the Ph.D. degree in electrical engineering at the Oregon State University, Corvallis. Mr. Hanumolu received Analog Devices Outstanding student designer award in 2002 and the Intel Ph.D. fellowship in 2004. He worked at Cypress Semiconductors and Intel research labs on various analog-mixed signal circuits.

This talk focuses on two aspects of transparent electronics, SnO2 transparent thin-film transistors (TTFTs) and transparent circuits. Both depletion- and enhancement-mode SnO 2 TTFTs are realized. The maximum effective mobility for the depletion- and enhancement-mode devices are 2 cm^2V^-1 s^-1 and 0.8 cm^2V^-1s^-1, respectively. The second part of this talk focuses on the fabrication procedure and the electrical characteristics of transparent circuits, which include inverters and ring oscillators. These circuits are highly transparent, exhibiting 75% optical transmittance in the visible portion of the electromagnetic spectrum, and are fabricated using indium gallium oxide as the active channel material and standard photolithography techniques. The n-channel indium gallium oxide thin- film transistors exhibit a peak incremental mobility of 7 cm^2V^-1s^-1, and a turn-on voltage of 2 V. A five-stage ring oscillator circuit (which does not employ level-shifting) is fabricated, and exhibits an oscillation frequency of 2.2 kHz with the gate and drain of the load transistor biased at 30 V. The maximum oscillation frequency observed is 9.5 kHz, with the gate and drain of the load transistor biased at 80 V.

Biography

Rick E. Presley grew up in Lebanon, OR, and received the B.S. degree in computer engineering from Oregon State University in 2001. He has been a graduate research assistant under Doctor John F. Wager at Oregon State University working on transparent electronics since 2003.

The traditional approach to improve the performance of critical building blocks has been the fine tuning of conventional architectures together with the advantages provided by technology scaling. However, fundamental limitations have arisen due to the technology scaling which require innovative system level techniques instead of the traditional tweaking of analog circuits. Moreover, all digital architectures do not provide a solution to these limitations either since analog to digital and digital to analog interfaces are unavoidable building blocks. This presentation shows that the strategic coupling between analog and digital systems combined with powerful signal processing tools is the most solid and open-minded approach for pushing the performance envelop of critical systems. In particular, the following two new analog to digital converters schemes will be presented: Analog to Digital Conversion via Signal Expansion: Analog to digital conversion is one of the key elements that has enabled the development and implementation of digital signal processing systems. However, the progress in the development of ADCs with higher speeds and resolutions is facing serious technological barriers. I believe that the further development of ADCs need advance signal processing techniques that couple analog and digital parts in an optimal fashion. At this end, I have proposed during my Ph.D. dissertation an ADC approach based on the quantization of coefficients obtained via the projection of a continuous-time signal over a set of basis functions. This ADC framework is motivated by the sampling of an input signal in domains which may lead to significantly less demanding A/D conversion characteristics; lower sampling rates and lower bit resolution requirements, which in principle can lead to lower power consumption. The proposed system efficiently parallelizes the ADC, which lowers the sampling rate requirements by increasing the number of basis functions on which the continuous time signal is projected, leading to a trade-off between sampling rate reduction and system complexity This research takes into consideration critical implementation issues in A/D converters providing the possibility of circuit level implementation of the proposed technique. A multicarrier receiver that couples the digital baseband with the signal expansion ADC front-end will be explained in detail. Digital Background Calibration Techniques of ADCs: Digital calibration techniques arise as a potential solution to the multiple impairments of analog circuits. These digital techniques are supported by their inherent robustness and flexibility. Additionally, the very large integration of digital circuits allows the implementation of complex adaptive calibration techniques whose foundations have been solidified by decades of research in signal processing. As an example of this, during my postdoctoral research project, I worked on the implementation of the Least-Mean-Square (LMS) algorithm for the calibration of a pipeline ADC. The pipelined ADC when is clocked at high speeds degrades its performance due mainly to device mismatches. It can be shown that these impairments can be overcome by passing the corrupted digital data through a linear filter. The optimum weights are found adaptively using the LMS algorithm. The training data used in the adaptation process is provided by a high resolution sigma-delta converter running at a fraction of the Nyquist sampling rate. A chip on 0.13 um technology has been taped-out to prove these ideas and although the silicon is not back yet for testing, simulations and circuit details will be discussed.

Biography

Sebastian Hoyos received the B.S. degree from the Pontificia Universidad Javeriana (PUJ), Bogota, Colombia in 2000, and the M.S. (2002) and Ph.D. (2004) degrees from the University of Delaware, Newark DE, all in electrical engineering. In the Fall of 2004, he enrolled in the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley, where he is a postdoctoral researcher at the Berkeley Wireless Research Center. He has carried out industrial consulting with Conexant Systems Inc., Red Bank, NJ. His research interests include communication systems, wireless communications, sensor network processing, robust signal processing, and mixed-signal high-speed processing and circuit design.

Wireless multihop networks are gaining popularity in applications such as mesh networks, wireless backbones, and sensor networks. Typically, communication protocols designed for such networks have explicitly or implicitly assumed an omnidirectional antenna at the radio layer. With recent progress in signal processing and antenna technologies, smart beamforming antennas have become feasible at cheaper prices, making them an attractive replacement to omnidirectional antennas. With beamforming antennas, the ability to guide RF energy in desired directions can lead to higher spatial reuse of the channel. Also, the higher communication range of beamforming antennas can be exploited to achieve stronger network connectivity, and to obtain fewer hop routes. However, existing protocols considered suitable for omnidirectional antennas are incapable of achieving these benefits. We have identified several weaknesses in these protocols, including new kinds of hidden terminal problems, "deafness", "MAC-layer capture", etc., that together degrade overall network performance. We have addressed each of these weaknesses systematically, and developed a better understanding of the theoretical improvements in capacity, achievable with beamforming antennas. Based on the insights, we have designed medium access control (MAC) and routing protocols to fully exploit the benefits of beamforming. We have demonstrated part of our solutions on a prototype testbed, equipped with electronically steerable antennas. This presentation will discuss our work on exploiting current and future smart-antenna systems in wireless multihop networks. I will describe the new challenges that arise with beamforming, and focus on the design, analysis, and evaluation of a new MAC and routing protocol. I will also summarize practical experiences from building a prototype testbed that utilizes beamforming antennas for multihop communication.

Biography

Romit Roy Choudhury is a Ph.D candidate of Computer Science at the University of Illinois at Urbana-Champaign. His research interests are in networking and systems, with an emphasis on wireless mesh networks, sensor networks, and ad hoc networks. Romit won the best paper award at the 2003 Personal Wireless Communications conference for his work on routing using directional antennas. He is a recipient of the Motorola Center for Communications Fellowship during 2003-2005, and the Vodafone Fellowship during 2005-2006. Further information is available at http://www.crhc.uiuc.edu/~croy/.

Syndromic surveillance is a field that involves the collection and analysis of data that precede diagnosis. The goal of syndromic surveillance is the early detection of both naturally occurring and bioterrorist-induced disease outbreaks. The scenario that is of greatest concern to syndromic surveillance systems is a large-scale release of inhalational anthrax. Data mining algorithms play a critical role in analysis of syndromic surveillance data by detecting anomalies that correspond to disease outbreaks. This talk will present the Population-wide Anomaly Detection and Assessment (PANDA) algorithm, which is specifically designed to monitor health-care data for the onset of an outbreak caused by an outdoor, airborne release of inhalational anthrax. At the heart of the PANDA algorithm is a causal Bayesian network which models the spatio-temporal effects of the outbreak on a population. The most unique aspect of the PANDA algorithm is its attempt to model every individual in the population under surveillance.

Biography

Weng-Keen Wong is an Assistant Professor of Computer Science at Oregon State University. He received his Ph.D. (2004) and M.S.(2001) in Computer Science at Carnegie Mellon University, and his B.Sc. (1997) from the University of British Columbia. His research areas are in data mining and machine learning, with specific interests in anomaly detection, surveillance algorithms, and mining large scale datasets. His Ph.D. thesis was entitled "Data Mining Algorithms for the Early Detection of Disease Outbreaks" and he is involved in the field of disease outbreak surveillance.

The goal of the Center for Wireless Integrated Micro-Systems at the University of Michigan is to develop functional micro-systems combining micro-power integrated circuits, wireless interfaces, and Micro-Electro-Mechanical Systems (MEMS) in a platform suitable for a wide range of applications. One example of such an application is a miniature integrated environmental monitoring system that uses an integrated wireless interface to transmit the recorded data using radio frequency signals. Such systems require electrically small antennas that can easily be integrated with their RF transceiver circuits and provide adequate bandwidth for the system to operate. Developing viable methods for designing such antennas is the main subject of this presentation. The presentation will begin with an overview of a number of antenna miniaturization techniques and examples of miniaturized antennas that are 15 to 20 times smaller than their commercial counterparts will be presented. The focus of the second half of the presentation will be on developing methods for bandwidth enhancement of such antennas. In particular, a method for designing electronically reconfigurable antennas will be studied in detail. The application of this technique in the design of an antenna with two operating bands that can be tuned independently over a wide frequency range will also be discussed and measurement results of this highly tunable dual-band antenna will be presented.

Biography

Nader Behdad received the Bachelor of Science degree in Electrical Engineering from Sharif University of Technology in September 2000 and the Master of Science degree in Electrical Engineering from The University of Michigan, Ann Arbor, in 2003. He is currently working towards the Ph.D. degree at the Department of Electrical Engineering and Computer Science of the University of Michigan, Ann Arbor. Mr. Behdad is the recipient of the best student paper award in the Antenna Applications Symposium held in Monticelo, IL, in September 2003, winner of the second prize in the student paper competition of the USNC/URSI National Radio Science Meeting in Boulder, CO, in January 2004, and the recipient of the Horace H. Rackham Predoctoral Fellowship from the Rackham School of Graduate Studies of The University of Michigan in 2005.

This talk will first present an overview of recent research activities in multimedia communication and wireless sensor networks in the Multimedia Communications Lab at the Wireless Center of Excellence, Florida Institute of Technology. Then, this talk will focus on research projects in the areas of wireless video and wireless sensor networks. In wireless video, two different applications will be presented: (1) video transmission over mobile wireless channel and MIMO systems, and (2) DSP and software implementation of video codec for mobile devices. In particular, in-depth discussion in video over MIMO channels will be given. In wireless sensor networks, two different applications will be presented: (1) chain-type wireless sensor network for infrastructure and environmental monitoring, and (2) energy efficient imaging sensor applications. In depth discussion in chain-type wireless sensor networks will be given. Experimental results for these applications will be shown to demonstrate that the proposed approaches are able to consistently outperform the existing schemes in terms of improved video quality, reduced computational complexity, and superior energy efficiency.

Biography

Chang Wen Chen is currently Allen Henry Endow Chair Professor and Director of the Wireless Center of Excellence at Florida Institute of Technology. He was on the faculty of Electrical Engineering Department at the University of Rochester from 1992 to 1996, on the faculty of Electrical and Computer Engineering Department at the University of Missouri-Columbia from 1996 to 2003. He also served as the Head of Interactive Media Group at David Sarnoff Research Labs in Princeton from 2000 to 2002. Currently, he is serving as the Editor-in-Chief for IEEE Trans. Circuits and Systems for Video Technology. He has been an Associate Editor for IEEE Trans. Circuits and Systems for Video Technology 1997-2005, an Associate Editor for IEEE Trans. Multimedia 2002-2005, and an Editor for IEEE MutliMedia Magazine 2003-2005. He is also on the Editorial Board of Journal of Visual Communication and Image Representation since 2000. He is serving as Technical Program Committee Chair for 2006 IEEE International Conference on Multimedia and Expo to be held in July 2006 in Toronto, Canada. He is also serving as Program Committee Chair for SPIE Conference on Multimedia Systems and Applications to be held in October 2005 in Boston, MA and for SPIE Conference on Multimedia on Mobile Devices in January 2006 in San Jose, CA. His research interests include image and video coding, joint source and channel coding, wireless and Internet video, wireless sensor networks, and multimedia communication and networking. His research is supported by NSF, DARPA, Air Force, NASA, Whitaker Foundation, and Kodak. He was elected a Fellow of IEEE in 2004 for his contributions in digital image and video processing, analysis, and communications. He received his BS from University of Science and Technology of China in 1983, MSEE from University of Southern California in 1986, and Ph.D. from University of Illinois at Urbana-Champaign in 1992.

Ecosystem Informatics is an interdisciplinary field that examines the prospects for advancing computer science and information technology research by focusing on the complex and often unique challenges found in ecosystem domains. In the first part of this talk, I will present an overview of this emerging field and the critical challenges that data mining faces in this field. In the second part, I will present my research on unsupervised pattern discovery for two environmental science problems. For the first problem, clustering remote sensing land cover data, I will present an ensemble based clustering technique, which provides a flexible and reliable solution to the high dimensionality problem we face. For the second problem, correlation pattern analysis of vegetation-precipitation data, I will introduce a novel approach to learning mixtures of local linear correlation models that is capable of finding nonlinear correlation patterns, and patterns that are only locally valid in the data.

Biography

Dr. Xiaoli Fern is an Assistant Professor of Computer Science at Oregon State University. She received her Ph.D (2005) in Computer Engineering from Purdue University and her M.S.(2000) and B.S.(1997) degrees from Shanghai Jiao Tong University. Her research interests are in machine learning and data mining, specifically in the area of unsupervised learning, including clustering, correlation analysis, dimension reduction, outlier detection and frequent pattern mining, etc. She is particularly interested in working with ecological and environmental data.

Flow control, including congestion control for data and rate control for multimedia, is an important issue in information transmission over wireless networks. In this talk, we formulate the problem as a concave optimization one and then we proposed a solution to the problem that can be modeled as a nonlinear singular perturbed system, and interpreted as dynamically adjusting both flow rates and number of connections of users. This modeling generates new insights to practice. We further explore the uniqueness of equilibrium rate and its global exponential stability, from both optimization and control perspectives. These results address stability, scalability and incremental deployment concerns from networking's point of view. Following theoretical insights, we design practical end-to-end schemes addressing the flow control problem in practice. Their performances are evaluated and characterized using both NS-2 simulations and actual experiments over Verizon Wireless 1xRTT commercial data network.

Biography

Minghua Chen is a Ph.D. candidate in Department of Electrical Engineering and Computer Sciences, University of California at Berkeley. He is a memeber of Video and Image Processing Lab, working with Prof. Avideh Zakhor. His dissertation identifies and developes new solutions for flow control in wireless networks, whose performance is guaranteed in theory and evaluated in practice. His research interests are in general area of information transmission over wireless networks, including source coding, flow control, network routing and coding, and wireless communications.

This seminar covers various energy storage systems mainly flywheels for land/sea/space applications. The flywheel is the system component responsible for storing energy in kinetic form when spinning at high speed. When selecting the appropriate design configuration for the flywheel, hub performance metrics such as stored energy vs. flywheel weight, stored energy vs. volume, or weight vs. cost can influence the overall performance of the system. For each application a different set of performance criteria are critical. For example, for space applications energy-vs.-weight ratio often has the highest importance, while for the Earth-based applications energy-vs.-volume and energy-vs.-cost ratios tend to be more significant. Flywheels are now being designed with new advancements in rotating machinery including non-contact magnetic bearings and permanent magnet motors/generators. New powerful magnet materials and power electronics enable flywheels to effectively fill the niche of short duration, high cycle life applications where batteries and ultra capacitors are not usable. High-speed operation and high reliability requirements limit selection of motors/generators used in modern flywheels to brushless and windingless-rotor types. Among these, permanent magnet (PM) machines have the most advantages, including higher efficiency and smaller size when compared with other types of motors/generators of the same power rating. They also exhibit lower rotor losses and lower winding inductances, which make them more suitable for a vacuum operating environment and the rapid energy transfer in flywheel applications. Very low cogging torque and robust rotor construction with very low part count are additional arguments for using PM motor/generators in flywheel applications.

Biography

Prof. Hamid Toliyat received the B.S, degree from Sharif University of Technology, Tehran, Iran in 1982, the M.S. degree from West Virginia University, Morgantown, WV in 1986, and the Ph.D. degree from University of Wisconsin-Madison, Madison, WI in 1991, all in electrical engineering. Following receipt of the Ph.D. degree, he joined the faculty of Ferdowsi University of Mashhad, Mashhad, Iran as an Assistant Professor of Electrical Engineering. In March 1994 he joined the Department of Electrical Engineering, Texas A&M University where he is currently E.D. Brockett professor of electrical engineering. Dr. Toliyat has received the prestigious Cyrill Veinott Award in Electromechanical Energy Conversion from the IEEE Power Engineering Society in 2004, Outstanding Professor Award in 2005 from Texas A&M, TEES Fellow Award in 2004, Distinguished Teaching Award in 2003, E.D. Brockett Professorship Award in 2002, Eugene Webb Faculty Fellow Award in 2000, and Texas A&M Select Young Investigator Award in 1999 from Texas A&M University. He has also received the Space Act Award from NASA in 1999, and the Schlumberger Foundation Technical Awards in 2001 and 2000. Dr. Toliyat is an Editor of IEEE Transactions on Energy Conversion, and was an associate editor of IEEE Transactions on Power Electronics. He is also Chairman of IEEE-IAS Electric Machines Committee, and is a member of Sigma Xi. He is a senior member of the Power Engineering, Industrial Applications, Industrial Electronics, Power Electronics Societies of the IEEE, and the recipient of the 1996 IEEE Power Engineering Society Prize Paper Award for his paper on the Analysis of Concentrated Winding Induction Machines for Adjustable Speed Drive Applications-Experimental Results. His main research interests and experience include analysis and design of electrical machines, variable speed drives for traction and propulsion applications, fault diagnosis of electric machinery, and sensorless variable speed drives. Prof. Toliyat has supervised more than 35 graduate students, published over 265 technical papers, raised over $3.4M in research funding, presented more than 35 invited lectures all over the world, and has 10 issued and pending US patents in these fields. He is the author of DSP-Based Electromechanical Motion Control, CRC Press, 2003, and the Co-Editor of Handbook of Electric Motors - 2nd Edition, Marcel Dekker, 2004. He was the General Chair of the 2005 IEEE International Electric Machines and Drives Conference in San Antonio, Texas.

Andlog-to-digital converters are often used in such instrumentation and measurement applications as weight scales, humidity or temperature or pressure sensors. These ADCs usually do not need high speed, but require very high absolute accuracy and linearity, as well as very low offset and and noise. Low power is also often an important design consideration. Ordinary delta-sigma ADCs are not well suited for such applications; however, a modified version, the incremental data converter, performs very well. In this seminar, the operating principles of incremental ADCs will be explained, and a recently fabricated 22-bit ADC with 0.3 ppm noise, 2 uV offset, 2 ppm gain error, and 4 ppm INL will be briefly described.

Biography

Gabor Temes received his undergraduate education in Hungary. He received his Ph.D. at the University of Ottawa in 1961. He is now a Professor in the School of Electrical Engineering and Computer Science at Oregon State University, and a Professor Emeritus of UCLA. He held positions at UCLA, Ampex Corp., Stanford University and BNR. He is a Life Fellow of IEEE. He received the Technical Achievement Award and the Education Award of the IEEE CAS Society, as well as the IEEE Centennial Medal. He is the recipient of the 1998 IEEE Graduate Teaching Award and received the IEEE Millennium Medal and the IEEE-CAS Golden Jubilee Medal in 2000. He was the recipient of the 2006 IEEE Gustav Robert Kirchhoff Award. Prof. Temes has written many books and papers on discrete and integrated circuit design.

Major challenges in realizing reliable and resource-efficient network services for wireless networks in the absence of infrastructure, are tentatively being addressed by the development of collaborative, cross-layer protocols. At present, security in wireless network designs is added as an after thought overlay, an approach that has migrated from our experiences in infrastructure-based wired networks. In this talk, I advocate that reliability, resource-efficiency and layer transparency cannot be considered disjointedly from network security, when wireless networks operate untethered in hostile environments. Network nodes must be able to verify that the information to be processed comes from a trustable source (authenticity), has not been altered in transit (integrity), is fresh (freshness), is local (locality verification) and restricted to authorized parties (confidentiality). Providing effective security in resource-constrained environments that are vulnerable to side-channel attacks requires a multi-modal approach in which cryptography is only one primitive/building block. I show the need for multi-modality by illustrating a series of vulnerabilities on location estimation methods proposed for wireless sensor networks in non-adversarial setting, and present a distributed passive solution for achieving resource-efficient secure location estimation. I also present a cross-layer design for resource-efficient group communications in wireless ad hoc networks.

Biography

Loukas Lazos received his Diploma in Engineering Degree in Electrical and Computer Engineering from the National Technical University of Athens, Greece, in 2000. In 2001, he joined the Network Security Lab at the Electrical Engineering Department at the University of Washington. He is a Ph.D. candidate expected to graduate the summer of 2006. His research interest include identifying, analyzing and modeling security threats in wireless ad hoc and sensor networks, as well as designing and analytically evaluating resource-efficient security protocols.

A typical wireless sensor network performs only one action: sensing the environment. Our requirement for intelligent interaction with the environment has led to the emergence of Wireless Sensor and Actor Networks (WSANs), where a group of sensors, actors and a central coordination entity (sink) linked by wireless medium perform distributed sensing and acting tasks. The evolution from WSNs, which can be thought of to perform only read operations, to WSANs, which can perform both read and write operations, introduces unique and new challenges that need to be addressed. In this presentation, I address one such challenge called "hazards", which is the out-of-order execution of queries and commands by due to a lack of coordination between sensors, actors and the sink. I identify three types of hazards and show with an example application, the undesirable consequences of these hazards. I also enumerate the associated challenges in a WSAN environment while addressing hazards. In this context, I discuss the basic design needed to address this problem efficiently. I propose a distributed and fully localized hazard-free approach that addresses the problem and the associated challenges based on the design. Through analytical studies, simulations and a prototype implementation, I study the performance of the proposed solution and two baseline strategies, and show that the proposed solution is efficient for a variety of network conditions.

Biography

Ramanuja Vedantham is a Ph.D. Candidate at The School of Electrical and Computer Engineering at Georgia Institute of Technology. His research interests are in the areas of wireless sensor and actor networks, and network protocols for multi-hop wireless networks. He received his M.S. in Computer Science from The University of Texas at Austin in 2002, and B.Tech. in Electrical Engineering from Indian Institute of Technology, Madras in 2000, where he was the recipient of several scholarships and awards. He has publications in ELSEVIER AD-HOC NETWORKS 2006, ELSEVIER COMPUTER COMMUNICATIONS 2006, BROADNETS 2005, ACM MobiHoc 2004, ICC 2005 and WICON 2005 among otherleading journals and conferences. He has reviewed several papers in leading conferences and journals (MOBICOM 2003-2006, INFOCOM 2004-2006, MOBIHOC 2004-2006, ToMC 2004-2006, ToN 2004-2006), and is a member of IEEE Communications society.

Circuit design for digital links and optical transceivers involves three critical challenges: wide bandwidth (BW), high gain and low power. Conventional single-stage approaches trade off wide BW for low gain because the gain-BW product increases as gain decreases; multiple-stage cascades trade off high gain for power, area and BW shrinkage. Peaking techniques for single-stage amplifiers achieve high gain simultaneously with high BW extension ratios (BWER), which means fewer stages. Increases in BWER and gain are achieved using capacitor-splitting and magnetic-coupling to sequence charging currents in bridged-shunt series and asymmetric T-coil amplifiers. UWB communication systems use a 3.1-10.6GHz spectrum. LNA design is critical in a UWB receiver; it should exhibit low return loss, low noise figure, high gain across 7.5GHz, and consume minimum power and die area. Cost and SoC considerations dictate the use of CMOS. Previous designs use common-source or distributed amplifiers; good performance is achieved, but reductions in power and die area are desired. A common-gate UWB LNA is described with low power and an area efficient impedance match along with stagger-compensated series peaking for BW extension.

Biography

David J. Allstot received the B.S. from Univ. of Portland, 1969, M.S. from Oregon State Univ., 1974, and Ph.D. from Univ. of California, Berkeley, 1979. He is the Boeing-Egtvedt Chair Professor of Engineering at the Univ. of Washington and Chair of the Dept. of Electrical Engineering. Dr. Allstot has advised more than 80 M.S. and Ph.D. graduates and published 250 papers. Awards include: IEEE Baker Award, IEEE Circuits and Systems Society Darlington Award, IEEE Intl. Solid-State Circuits Conference B. Winner Award, IEEE Circuits and Systems Society Technical Achievement Award, and Aristotle Award, Semiconductor Research Corp.. His service includes: Associate Editor, IEEE Trans. on Circuits and Systems, 1990-1993, and Editor, 1993-1995; Board of Governors, IEEE Circuits and Systems Society, 1992-1995; Technical Program Committee, IEEE Intl. Solid-State Circuits Conference, 1994-2004; Executive Committee Short Course Chair, IEEE Intl. Solid-State Circuits Conference, 1996-2000; Distinguished Lecturer, IEEE Circuits and Systems Society, 2000-2001; Co-General Chair, IEEE Intl. Symposium on Circuits and Systems, 2002. He has served as Distinguished Lecturer for the CAS Society and is currently (2006-2007) Distinguished Lecturer for the Solid-State Circuits Society. He is an IEEE Fellow.

Orthogonal frequency division multiplexing (OFDM) is a popular modulation technique for wireless digital communications. It provides a relatively straightforward way to accommodate high data rate links over harsh wireless channels characterized by severe multipath fading. OFDM has two primary drawbacks, however. The first is a high sensitivity to time variations in the channel caused by Doppler, carrier frequency offsets, and phase noise. The second, and the focus of this talk, is that the OFDM waveform has high amplitude fluctuations, a drawback known as the peak-to-average power ratio (PAPR) problem. The high PAPR makes OFDM sensitive to nonlinear distortion caused by the transmit power amplifier (PA). Without sufficient power backoff, the system suffers from spectral broadening, intermodulation distortion, and, consequently, performance degradation. High levels of backoff reduce the efficiency of the PA. For systems with limited power resources, such as mobile battery-powered devices, this problem is particularly detrimental. A new PAPR mitigation technique is presented. In constant envelope OFDM (CE-OFDM), the high PAPR OFDM signal is transformed to a constant envelope 0 dB PAPR waveform by way of angle modulation. The constant envelope signal can be efficiently amplified with nonlinear power amplifiers thus achieving greater power efficiency. In this talk, the fundamental aspects of the CE-OFDM modulation are studied, including the signal spectrum, the signal space, optimum performance, and the performance of a practical phase demodulator receiver. Performance is evaluated over a wide range of multipath fading channel models. It is shown that CE-OFDM outperforms conventional OFDM when taking into account the effects of the power amplifier. System details of a hardware implementation are also discussed.

Biography

Steve Thompson received the BSc degree from Arizona State University (1999), and the MSc and PhD degrees from the University of California, San Diego (2001 and 2005, respectively) all in Electrical Engineering. From 1997 to 1998 he worked as an Associate Engineer at Inter-Tel, Chandler, AZ. He also worked summer internships at Los Alamos National Laboratory, Los Alamos, NM (1998); and SPAWAR Systems Center, San Diego, CA (2001 and 2004). Dr. Thompson's research interests are in the general area of digital communications. Currently, he is working on constant envelope OFDM at UCSD as a Postdoctoral Scholar.

The widespread use of micro and nanofabrication technologies has resulted in the development of chips the size of a postage stamp, capable of rapidly processing biological samples as small as a single cell. The goal is to create fully automated systems that are on chips, easy to use—and general, universal, and inexpensive. This talk will first provide a general overview of the current biosensing technology by applying the principles of optical and electrical detection. It will then focus on the efforts of the Biomedical Microdevices and Nanotechnology laboratory at Portland State University where the goal is to develop a wide range of biosensors and bio-detectors for myriad applications, ranging from health care to environmental monitoring. A wide range of sensing platforms will be presented-the use of basic silicon micro electrode array technology for developing biological cell based sensors, use of nanostructured material for electrical detection of biomolecules, application of optical micro cavities as sensor platforms and the use of multilayered light emitting devices as biochemical sensors. The long term goal is to develop “system on a chip” devices suitable for implementation in the ambient atmosphere. The multi-disciplinary nature of the research will be highlighted; the integration of concepts of life sciences, physical sciences with engineering towards the development of functional platforms will be elucidated.

Biography

Shalini Prasad received her Bachelor’s degree in Electronics and Communication Engineering from University of Madras in 2000. She obtained her PhD in Electrical Engineering from University of California Riverside in 2004. Her research interests are multi-disciplinary and are in the emerging fields of Bio-Micro Electrical Mechanical Systems (Bio-MEMS) and nanotechnology. Her current research interests are directed towards developing scalable micro/nano platforms as sensors and detectors. Her work has been reported in number of peer reviewed journals and popular press. She is the recipient of the Association for Lab Automation Awards and Whitaker Award for Bio-MEMS research.

This talk will examine the concept of trustworthy computing from the perspective of the skeptical end-user. We will examine the concept of trust, and how this is, or is not being applied to trustworthy computing today. Specifically, we will examine why trustworthy computing fails to gain trust, and the roadblocks we face in transitioning what I call ‘faith-based’ computing.

Biography

Carlos Jensen is an Assistant Professor in the School of Electrical Engineering and Computer Science at the Oregon State University. His research is in the usability of privacy and security; developing new analysis, visualization, and interface techniques to build awareness and support users in making better decisions. As part of this work, Jensen studies users’ mental models and understanding of privacy and security, as well as the tradeoffs they make. Jensen is also leading an effort to index, analyze, and visualize website data-practices on a large scale in order to automatically classify sites, detect data-sharing networks, and evaluate the impact of regulation.

Wind energy is often installed in rural, remote areas characterized by weak, unbalanced power transmission grids. In induction wind generators, unbalanced three-phase stator voltages cause a number of problems, such as saturation, over-current, and stress on the mechanical components from torque pulsations. Therefore, for high levels of unbalance, induction wind generators are switched out of the network. This can further weaken the grid. A special type of induction generator, called a doubly-fed induction generator, has found wide use in high-power wind applications. The wound rotor construction of a doubly-fed induction machine allows for rotor current control via a rotor-connected power converter. Control of the rotor currents in turn allows for variable speed operation and reactive power control. The goal of the research is to further extend the control strategy beyond the standard variable speed and reactive power control to also include compensatory features that can correct for the problems caused by an unbalanced stator voltage. This improves the robustness and all-around performance of doubly-fed induction wind generators, allowing them to operate in conditions in which they would normally be removed from the grid. Unbalanced stator voltages appear as a positive and negative sequence 3-phase voltage at the stator terminals. The negative sequence causes a flux field in the machine to rotate counter to the main, positive sequence direction. This causes a second harmonic pulsation in the machine torque (active power) and reactive power. Also, the machine stator currents become unbalanced. When using synchronous frame vector control, the second harmonic disturbance caused by the unbalance will be propagated through all synchronous frame variables. The developed control compensates for the second harmonic, and hence the unbalance, by injecting a supplementary rotor voltage to cancel the second harmonic in both the torque and stator reactive power synchronous frame control loops. This also has the effect of reducing the stator current unbalance. This novel control was implemented and tested on a 15 kW doubly-fed induction machine. Theory and results will be presented.

Biography

Ted Brekken is postdoctoral researcher at the University of Minnesota and a laboratory instructor at the University of St. Thomas in St. Paul Minnesota. He received his B.S., M.S., and Ph.D. from the University of Minnesota in 1999, 2002, and 2005 respectively. He studied electric vehicle motor design at Postech in Pohang, South Korea in 1999. He also studied wind turbine control at the Norwegian University of Science and Technology in Trondheim, Norway in 2004-2005 on a Fulbright scholarship. His research interests include control, power electronics and electric drives; specifically digital control techniques applied to renewable energy systems.

Intel is an industry leader in Chip technology and manufacturing. Given the acceleration of digital convergence and ever increasing need for higher performance processors at optimal cost and power efficiency, it is important than ever before to innovate and deliver leading edge manufacturing technologies and products in a cost effective manner. Intel's Automated Manufacturing Technology is a critical ingredient enabling Intel's technology leadership. This talk will characterize the chip industry environment today, key drivers for Automated Manufacturing technology (AMT) at Intel and walk-thru examples of AMT.

Biography

Chandra Mouli is Director of Automated Manufacturing Technology in Intel’s Logic Technology Development (LTD) group. Mouli joined Intel in early 1998 initially focusing on developing Computer Simulation models and techniques to streamline and optimize manufacturing process flow. Between 1989 and 1995, he worked in the California Technology Development group where he led numerous fab data automation initiatives. From 2001 through 2005, he also led the migration of the core Manufacturing Execution System (MES) to next generation architecture that is widely recognized as industry leading. He invented the patent pending middleware architecture and framework that serves as the backbone of the manufacturing workflow and automated decision making in Intel fabs. He has numerous publications and 7 patents pending. Chandra is responsible for defining computing architecture directions and roadmap for Intel. He is also the technical spokesperson for Intel’s Automated Manufacturing Technology to the press and industry.

This presentation illustrates the new electrical energy revolution - the twenty-first century revolution that is completely changing the original twentieth century electrification revolution. With present concerns about energy costs and depletion of fossil fuels, technologies that support sustainable energy processes are becoming critical. It is shown how that power electronics, the processing of electrical energy with electronic circuits, is the driver behind this revolution. Modern portable devices, new energy resources, the fundamental changes occurring in automobiles, and many other areas have power electronics as a critical enabling technology. These innovations are the basis for alternative and sustainable energy technologies for the future. The tasks require nonlinear switching circuits, the core technology of power electronics. Examples include electronic motor drives, dc-dc converters, solar energy power conditioning circuits, "digital power," electric and hybrid automobiles, single-chip power converters, high-efficiency power supplies, special semiconductor devices, and even applications to audio amplifiers. Emphasis is on the urgent need for engineers familiar with the challenges and design tools of power electronics to help move into a new energy era.

Biography

Philip T. Krein is the Grainger Endowed Director's Chair Professor of Electric Machinery and Electromechanics at the University of Illinois. His research interests address all aspects of power electronics, machines, and drives, with emphasis on nonlinear control approaches. He is the author of an undergraduate textbook on power electronics. He has been a senior Fulbright Scholar in the United Kingdom and is a past President of the IEEE Power Electronics Society. In 2003 he received the IEEE William E. Newell Award in Power Electronics, and at present is a Distinguished Lecturer for the IEEE Power Electronics Society.

An overview of the basic physics theory required to build compact MOSFET models and a unified treatment of inversion-charge and surface-potential models are provided. The needs of digital, analog and RF designers as regards the availability of simple equations for circuit design are taken into account. Compact expressions for hand analysis or for automatic synthesis, valid in all the operating regions (weak-moderate-strong inversion) are presented. Since designers in advanced technologies are increasingly concerned with fluctuations, the modeling of fluctuations is emphasized. A unified approach for both space (matching) and time (noise) fluctuations is introduced.

Biography

Carlos Galup-Montoro is a professor in the Electrical Engineering Department at the Federal University of Santa Catarina, Brazil. His research interests include semiconductor-device modeling and mixed-ICdesign. Galup- Montoro has an engineering degree in electronics and a doctor of engineering degree from the Institut National Polytechnique de Grenoble, France.

David Chassin will be present an overview of current work at PNNL in the energy sector. David will discuss electric grid research, including high performance computing as it relates to electric system modeling, planning, and operations, the role of buildings in grid operations, and the new Integrated Energy Operations Center.

Biography

David Chassin is a staff scientist at PNNL where he has working since 1992 on building systems modeling, diagnostic, and performance. He recently led the Data Intensive Computing focus area for the Laboratory's Computing Initiative. His past work includes investigations in the behavior of complex adaptive systems and the behavior of autonomous agent systems.

Wireless resource allocation is an important strategy to combat detrimental effects of wireless channels, optimize the allocations of limited resources, and control the interferences, so as to provide the desired services and optimize the system performances. To reduce the overhead and signaling imposed by such strategy, the mobiles of the next generation networks have their own autonomies for resource allocation in a distributive way. However, users’ non-cooperative competition of the radio resources results in low system efficiency. So how to ensure cooperation among autonomous and distributed users is one of the most important wireless networking research topics. In this talk, we propose a general game theoretical framework to enforce cooperation by the following approaches: incentive-based, referee-based, punishment-based, and relay/collaborative-communication-based. For incentive-based approach, different from the traditional “Pricing Anarchy”, we propose a bargaining method to have mutual benefits. Specifically, a fair and simple scheme to allocate subcarrier, rate, and power for multiuser single cell 4G/WLAN OFDMA systems is considered. The problem is to maximize the overall system rate, under each user's maximal power and minimal rate constraints, while considering the fairness among users. The approach proposes the fairness and low complexity implementation based on Nash Bargaining Solutions and Coalitions. We also briefly mention the other approaches. For referee based-approach, a virtual mediator is introduced to improve the outcome of the non-cooperative game, which is employed to multi-cell WMAN/OFDMA networks. For relay/collaborative-communication-based approach, a buyer/seller game such as Stackelberg game is studied to have distributive resource allocation over this new communication paradigm. Finally, for the punishment-based approach, a repeated game with threat of future punishment is proposed to suppress users' greediness and enforce cooperation in Ad Hoc networks.

Biography

Dr. Zhu Han received his BS in Electronic Engineering from Tsinghua University, China, 1997, received MS in 1999, and Ph.D. on Nov. 2003 from Electrical and Computer Engineering Department of University of Maryland, College Park. From 1997 to 2000, he was a Graduate Research Assistant at the University of Maryland. From 2000 to 2002, he was an Engineer in the R&D Group of ACTERNA, LLC Maryland. He is currently a Research Associate at the University of Maryland. His research interests include wireless resource allocation and management, wireless communications and networking, game theory, and wireless multimedia. Dr. Han is a member of the Technical Programming Committee for the IEEE ICC of 2004, 2005, and 2007, the IEEE VTC, Spring 2004, the IEEE CCNC 2005, the IEEE Globecom 2005 and 2006, and the IEEE WCNC 2005, 2006, 2007.

Internet megatrends predict the emergence of a new wave of end user programming of both web applications, services and components. In this talk I will present evidence for these trends, demonstrate the ADIEU end user programming environment, and discuss the emerging technical and social issues generated by these developments.

Biography

Sam S. Adams is an IBM Distinguished Engineer within IBM's Research Division. Mr. Adams joined the IBM Consulting Group in 1994 as a founding member of IBM's Object Technology Practice. In 1995 he helped create the IBM Object Foundry, which was the genesis of IBM's world-wide software reuse infrastructure. Sam was elected that year to the IBM Academy of Technology where he led an Academy study on Self-configuring systems, the initial IBM effort on what was to become Project eLiza and Autonomic Computing in 1999. In 1996, Sam was named one of IBM's first Distinguished Engineers. He spent 1997 in IBM Research investigating adaptive systems technology and then joined Network Computing Software Division in 1998 to help drive emerging technologies into IBM products and services. Sam was IBM's technical architect and strategist for XML in 1999 and helped create the concept of Service Oriented Architectures, which forms the basis for today's Web Services and Grid Services efforts. He returned to IBM Research in 2000, spending the next 2 years on exploratory research into semantic processing and common sense computing. As part of IBM's On Demand Computing effort, Sam is currently focusing on empowering end users to develop their own web application via a radically simplified approach to programming web services. A Native American of Cherokee and Sac-and-Fox descent, Sam is IBM's Executive liaison to AISES, the American Indian Science and Engineering Society. Prior to joining IBM, Sam spent 8 years as co-founder and Chief Scientist for Knowledge Systems Corporation in Cary, NC, a software and services startup that played a major role in the commercial acceptance of the Smalltalk programming environment.

One of the most common electrical engineering viewpoints in the 21st century has been the eventual end to transistor scaling (Moore’s Law) due to fundamental physical limitations. Much current research focus has recently touted the new age of nanotechnology, such as quantum computing, magnetics/spintronics, and MEMS. While these research areas certainly hold enormous potential, such technologies will succeed only if they are integrated together with the semiconductor processing goliath and its vast amount of computational ability—thereby taking full advantage of nanoelectronics promise. At the same time in parallel, Moore’s Law will continue to scale transistors, giving us unprecedented computing performance/benefits that we unknowingly take for granted. In this talk, I will describe in broad context how we can use the benefits of transistor scaling to effectively enhance the longevity and usefulness of such transistor scaling. I will first talk about my past and current research in designing new circuit architectures in deep submicron technologies that take advantage of CMOS’s benefits, namely ubiquitous transistors, increased transistor performance, and system-on-a-chip possibilities--while compensating for its growing disadvantages, such as process variation, inaccurate parasitic/device modeling, and increased design time/cost. I will then discuss how such research in CMOS circuits provides a natural interface with non-CMOS technologies, creating new and interesting interdisciplinary research areas. Curiosity, an open mind, and contrarian thought are especially welcome.

Biography

Patrick Chiang received the B.S. degree in electrical engineering and computer sciences from the University of California, Berkeley, in 1997, and the M.S. and Ph.D. degrees in electrical engineering from Stanford University in 2001 and 2006. In 1998, he was with Datapath Systems (now LSI Logic), working on analog front-ends for DSL chipsets. In 2002 he was a research intern at Velio Communications (now Rambus) working on 10GHz clock synthesis architectures. In 2004 he was a consultant at startup Telegent Systems, evaluating low phase noise VCOs for CMOS mobile TV tuners. In 2006 he was a visiting NSF postdoctoral researcher at Tsinghua University, China, investigating low power, low voltage RF transceivers. His interests are in the design and implementation of new architectures for mixed signal circuits in deep submicron CMOS. In high speed serial links, his interests include the design of low power, high data rate, parallel I/O's; and the implementation of 20+ GHz, 4-5 bit low power ADCs for limited bandwidth channels. In RF circuits, his research revolves around the design of reconfigurable, self-healing RF transceivers to compensate for process variation in future CMOS technologies.

In the last few years, there has been a big interest in Ad Hoc Wireless Networks as they have tremendous military and commercial potential. An Ad Hoc Wireless Network is a wireless network, comprising of mobile nodes (which can also serve as routers) that use wireless transmission, having no infrastructure (central administration such as a Base Station in a Cellular Wireless Network or an Access Point in a Wireless LAN). Ad Hoc Wireless Networks can be set up anywhere and anytime as they eliminate the complexities of infrastructure setup. Ad Hoc wireless Networks find applications in several areas. Some of these include: military applications (establishing communication among a group of soldiers for tactical operations as setting up of a fixed infrastructure in enemy territories or in inhospitable terrains may not be possible), collaborative and distributed computing, emergency operations, wireless mesh networks, wireless sensor networks, and hybrid (integrated Cellular and Ad hoc) wireless networks. In this talk, I first present a brief overview of the major issues that influence the design and performance of Ad Hoc Wireless Networks. As the performance of any wireless network hinges on the Medium Access Control (MAC) protocol, more so for Ad Hoc Wireless Networks, I present, in detail, three recently proposed MAC protocols for Ad Hoc Networks.

Biography

C. Siva Ram Murthy received the Ph.D. degree in Computer Science from the Indian Institute of Science, Bangalore in 1988. Since then he has been on the faculty of the Department of Computer Science and Engineering at the Indian Institute of Technology, Madras. He is currently a Professor with the same department since September 2000. He has held visiting positions at the German National Research Centre for Information Technology (GMD), Bonn, Germany, the University of Stuttgart, Germany, the University of Freiburg, Germany, the Swiss Federal Institute of Technology (EPFL), Switzerland, and the University of Washington, Seattle, USA. He is the co-author of the textbooks Resource Management in Real-time Systems and Networks, (MIT Press, Cambridge, Massachusetts, USA), WDM Optical Networks: Concepts, Design, and Algorithms, (Prentice Hall, Upper Saddle River, New Jersey, USA), and Ad Hoc Wireless Networks: Architectures and Protocols, (Prentice Hall, Upper Saddle River, New Jersey, USA). His research interests include parallel and distributed computing, real-time systems, lightwave networks, and wireless networks. He has published more than 120 international journal and 100 international conference papers in these areas. Dr.Murthy is a recipient of Best Ph.D. Thesis Award from the Indian Institute of Science, Indian National Science Academy (INSA) Medal for Young Scientists, and Dr. Vikram Sarabhai Research Award. He is a co-recipient of Best Paper Awards from the 5th IEEE International Workshop on Parallel and Distributed Real-Time Systems (WPDRTS), and the 6th and 11th IEEE Annual International Conference on High Performance Computing (HiPC). He is a Fellow of the Indian National Academy of Engineering, an Associate Editor of IEEE Transactions on Computers, and a Subject Area Editor of Journal of Parallel and Distributed Computing.

A significant untapped renewable energy source exists in the world’s oceans: it is estimated that if 0.2% of the oceans’ untapped energy could be harnessed, it could provide power sufficient for the entire world. This presentation discusses the opportunities for ocean wave power to become a new, reliable and clean source of affordable renewable energy. Wave energy research and developments at Oregon State University (OSU) will be presented, in addition to the developing opportunities for the State of Oregon to become a leader in wave power.

Biography

Ted K.A. Brekken is an Assistant Professor in Energy Systems at Oregon State University. He received his B.S., M.S., and Ph.D. from the University of Minnesota in 1999, 2002, and 2005 respectively. He studied electric vehicle motor design at Postech in Pohang, South Korea in 1999. He also studied wind turbine control at the Norwegian University of Science and Technology in Trondheim, Norway in 2004-2005 on a Fulbright scholarship. His research interests include control, power electronics and electric drives; specifically digital control techniques applied to renewable energy systems.

Object-oriented programming has worked quite well – so far. What are the objects, how do they relate to each other? Once we clarified these questions we typically feel confident to design and implement even the most complex systems. However, objects can deceive us. They can lure us into a false sense of understanding. The metaphor of objects can go too far by making us try to create objects that are too much inspired by the real world. This is a serious problem, as a resulting system may be significantly more complex than it would have to be, or worse, will not work at all. We postulate the notion of an antiobject as a kind of object that appears to essentially do the opposite of what we generally think the object should be doing. As a Gedankenexperiment antiobjects allow us to literally think outside the proverbial box or, in this case outside the object. I will discuss several applications of antiobjects in education and game design. In a soccer simulation example antiobjects are employed as part of a game AI called Collaborative Diffusion. In Collaborative-Diffusion based soccer the player and grass tile agents are antiobjects. Counter to the intuition of most programmers the grass tile agents, on top of which all the players are moving, are doing the vast majority of the computation, while the soccer player agents are doing almost no computation. This presentation illustrates that this role reversal is not only a different way to look at objects but, for instance, in the case with Collaborative Diffusion, is simple to implement, incremental in nature and more robust than traditional approaches.

Biography

Alexander Repenning is a professor of computer science at the University of Colorado and the founder of AgentSheets Inc. Repenning’s research interests include education, end-user programmable agents, and artificial intelligence. He has worked in research and development at Asea Brown Boveri, Xerox PARC, Apple Computer, and Hewlett Packard. Repenning is the creator of the AgentSheets simulation and game-authoring tool. He has taught game design nationally at Stanford, the MIT Media Lab, and University of Colorado as well as internationally in Europe and Japan. His work has received numerous awards including the Gold Medal from the mayor of Paris for “most innovative application in education of the World Wide Web” and “best of the best innovators” by ACM. Repenning is an advisor to the National Academy of Sciences, the European Commission, the National Science Foundation, The Japanese Ministry of Education and the Organisation for Economic Co-operation and Development.

CoreCalc is an implementation in C# of core spreadsheet functionality, intended as a testbed for implementation experiments. This talk will describe two such experiments. First, an experiment using runtime bytecode generation shows that formula recalculation in a safe managed code implementation can achieve speeds comparable with that of Excel, and far better than other open source spreadsheet programs, despite the C/C++ implementations of the latter. Joint work with Thomas Iversen. Second, we present the first implementation of sheet-defined functions as proposed by Blackwell, Burnett and Peyton Jones in their 2003 ICFP paper. A sheet-defined function can be recursive, can work on matrices and can be higher order. This is shown to provide considerable expressiveness and simplification on some examples from the life insurance industry. Joint work with Daniel Cortes and Morten Hansen. -- Version 0.5 of CoreCalc and several related technical reports can be found at http://www.itu.dk/people/sestoft/corecalc/

Biography

Peter Sestoft is professor of information technology and works mainly with programming language technology and software development. He is developer or co-developer of various open source software, including the Moscow ML implementation of Standard ML, and the C5 Generic Collection Library for C# and CLI, which is distributed with the Mono implementation of .NET. He is co-author of the standard text on partial evaluation (1993, with Jones and Gomard), and author of Java Precisely (2002 and 2005) and C# Precisely (2004 and 2006).

James Clerk Maxwell stands shoulder to shoulder with Newton and Einstein, yet even those of us who have spent decades working with Maxwell's equations are almost totally unfamiliar with his life and times. This presentation, from the viewpoint of a microwave engineer, draws on many sources in providing an understanding of James Maxwell himself. What was Maxwell like as an infant? What was the tragedy at eight years old that profoundly influenced his life? What unique means of transportation did young Maxwell use to escape a cruel tutor? What memorable event occurred on his first day of school? When did he publish his first papers, and what were they about? What did Maxwell have to do with the rings of Saturn? Why did he lose his job as a professor? Why did he have a hard time getting another job? What was his wife like? What is Maxwell's legacy to us? The answers to these questions provide insight into Maxwell the person and add an extra dimension to those four simple equations we have studied ever since. There are no equations in this presentation. The presentation is appropriate for anyone with a general interest in the origins of modern physics. For electronic handouts for the lecture, visit www.sonnetsoftware.com and click on the large Distinguished Microwave Lecture Series" button at the bottom of the "News" section.

Biography

James C. Rautio received a BSEE from Cornell in 1978, a MS Systems Engineering from University of Pennsylvania in 1982, and a Ph. D. in electrical engineering from Syracuse University in 1986. From 1978 to 1986, he worked for General Electric, first at the Valley Forge Space Division, then at the Syracuse Electronics Laboratory. At this time he developed microwave design and measurement software, and designed microwave circuits on Alumina and on GaAs. From 1986 to 1988, he was a visiting professor at Syracuse University and at Cornell. In 1988 he went full time with Sonnet Software, a company he had founded in 1983. In 1995, Sonnet was listed on the Inc. 500 list of the fastest growing privately held US companies, the first microwave software company ever to be so listed. Today, Sonnet is the leading vendor of 3-D planar high frequency electromagnetic analysis software. Dr. Rautio was elected a fellow of the IEEE in 2000 and received the IEEE MTT Microwave Application Award in 2001 and is an adjunct professor at Syracuse University.

Human vision is so powerful that we seldom give a second thought to our ability to immediately identify the objects in our surroundings. However, the problem of object recognition has proved very challenging for computer vision. A major source of the difficulty is the large range of variations in appearance that may occur, due to factors such as changes in 3D viewpoint, varying illumination, partial visibility, and background clutter. Fortunately, there has been rapid progress on this problem within the past few years through the use of an approach known as invariant local feature matching. Thousands of these local features can be extracted from an image to describe small overlapping regions, and each feature is designed to be invariant to a range of image transformations, such as changes in scale, orientation, brightness, and local deformations. Furthermore, the features are designed to be very distinctive, so that a single feature can be used to select a correct match from a large database. Fast methods for nearest-neighbor access in high-dimensional spaces allow large databases of features to be matched in real time. This talk will present an overview of the invariant feature approach, as well as some recent applications such as location recognition and automated stitching of digital images into panoramas.

Biography

David Lowe is a professor of Computer Science at the University of British Columbia and a Fellow of the Canadian Institute for Advanced Research. He received his Ph.D. in computer science from Stanford University in 1984. From 1984 to 1987 he was an Assistant Professor at the Courant Institute of Mathematical Sciences at New York University. He is a member of the scientific advisory board for Evolution Robotics. His research interests include object recognition, local invariant features for image matching, robot localization, and computational models of human visual recognition.

End User Programming has been one of the holy grails of software development even before Alan Kay coined the the term "Personal Computing". While there have been notable successes such as the spreadsheet and Apple's HyperCard, the vast majority of computer users are limited to software developed by professionals due to the complexity of the tools and lanugages currently available. But this situation is in the process of making a dramatic change for the better. The past several years have seen the large scale adoption of technology such as Wikis and Blogs, which have ushered in the era of participatory media, where anyone with internet access can create and publish information in a wide variety of forms. What is coming after this first wave is an expansion from participatory media to participatory software development, where users can finally development, deploy, share and reuse software applications and components made by themselves for themselves. This summer, Chris Anderson, the editor of WIRED magazine, published the first book on a concept he created in 2005, The Long Tail, which proposes a new economic model of mass market of niches that describes the success of internet giants such as Google, Amazon and eBay. This talk will discuss the future of end user programming from a Long Tail perspective, and why IBM is preparing for a large scale shift in the industry toward software development by the users for the users.

Biography

Sam S. Adams is an IBM Distinguished Engineer within IBM's Research Division. Mr. Adams joined the IBM Consulting Group in 1994 as a founding member of IBM's Object Technology Practice. In 1995 he helped create the IBM Object Foundry, which was the genesis of IBM's world-wide software reuse infrastructure. Sam was elected that year to the IBM Academy of Technology where he led an Academy study on Self-configuring systems, the initial IBM effort on what was to become Project eLiza and Autonomic Computing in 1999. In 1996, Sam was named one of IBM's first Distinguished Engineers. He spent 1997 in IBM Research investigating adaptive systems technology and then joined Network Computing Software Division in 1998 to help drive emerging technologies into IBM products and services. Sam was IBM's technical architect and strategist for XML in 1999 and helped create the concept of Service Oriented Architectures, which forms the basis for today's Web Services and Grid Services efforts. He returned to IBM Research in 2000, spending the next 2 years on exploratory research into semantic processing and common sense computing. As part of IBM's On Demand Computing effort, Sam is currently focusing on empowering end users to develop their own web application via a radically simplified approach to programming web services. A Native American of Cherokee and Sac-and-Fox descent, Sam is IBM's Executive liaison to AISES, the American Indian Science and Engineering Society. Prior to joining IBM, Sam spent 8 years as co-founder and Chief Scientist for Knowledge Systems Corporation in Cary, NC, a software and services startup that played a major role in the commercial acceptance of the Smalltalk programming environment.

Darcs is a distributed revision control system based on a formalism for manipulating changes, which allows for a system that is change-based rather than version-based. This talk will introduce this formalism. I will then show how we can use the type system of Haskell to verify the correctness of change-manipulation code.

Biography

David Roundy has just started as an assistant professor in the Physics Department here at OSU, working in the field of Condensed Matter Theory. He received his B.A. and Ph.D. at Berkeley and did postdocs at MIT and Cornell. He created darcs in 2002 after becoming interested in the problem of how to describe and manipulate changes. Around the same time, David learned the Haskell language, and hasn't been the same since. While not working on darcs or physics, David enjoys reading, knitting, crocheting and playing harmonica, but not all at the same time.

Programming is not easy even for trained professional programmers. Environments like spreadsheets allow end users to develop programs. It is not surprising that a majority of such end-user programs have non-trivial faults. Some surveys estimate that up to 90% of commercial spreadsheets have faults in them. Many of these faults have huge financial impact, and quite a few of these "horror stories" have been reported in the news media. The main cause of the high incidence of errors in end-user spreadsheets is the lack of tool support for error detection/prevention. Since it is not feasible to train all end users in "good" software engineering practices, how can the benefits of software engineering techniques be brought within the reach of end users? In this context, the main thrust of my research has been towards exploring ways to help end users develop safer spreadsheets. In the talk, I will present some of the tools we have developed and results from empirical evaluations of the tools.

Biography

Robin Abraham is a Ph.D. candidate at the School of Electrical Engineering and Computer Science, Oregon State University. He received his BTech. (Hons.) in Aerospace Engineering from the Indian Institute of Technology, Kharagpur, India in 1997. After working as a database administrator and software engineer for three years, he joined Oregon State University in 2000 and completed his masters in Computer Science in 2003. While at the Oregon State University, he was awarded the Best Graduate Teaching Assistant award in 2001, and the Best Graduate Research Assistant award in 2006 by the College of Engineering. Mr Robin Abraham and his advisor Dr Martin Erwig won the Best Paper Award at the IEEE Symposium on Visual Languages and Human- Centric Computing, 2004.

Through a series of examples, I will demonstrate presentation techniques that enhance or detract from the message you are trying to deliver. I will discuss how much information to put on a single slide, how slides should relate to one another, subliminal messages slides may convey, when to be silent, the use of gestures, question- answering strategies, and more.

Biography

Professor Michael J. Quinn earned an honors B.S. in mathematics from Gonzaga University in 1977, an M.S. in computer sciences from the University of Wisconsin-Madison in 1979, and a Ph.D. in computer science from Washington State University in 1983. From 1979 to 1981 he worked for Tektronix, Inc. as a software engineer. He was an assistant professor of computer science at the University of New Hampshire from 1983 to 1989, before joining Oregon State University in 1989. He served as interim Head of Computer Science in 1997-1998 and Head of Computer Science from 1998-2002. Dr. Quinn has authored or co-authored dozens of refereed publications and eight books in the areas of parallel computing, computer ethics, and computing concepts.

The next big thing in computer graphics for visualization, games, and special effects is GPU Programming, the ability to place your own code right on the graphics card. This talk will discuss where the code goes in the whole graphics process, how it is written, the math that goes with it, and what you can do with it. There will be lots of intriguing examples!

Biography

Mike Bailey is a Professor of Computer Science at Oregon State University. His areas of interest include scientific visualization, high performance computer graphics, GPU programming, and computer aided design and analysis. Mike received his PhD from Purdue University in Computer Graphics and Computer Aided Design. He has since worked at Sandia National Labs, Purdue University, Megatek Corporation, the San Diego SUpercomputer Center, and the University of California San Diego. Mike enjoys teaching and has taught a variety of college classes to over 3,200 students. UCSD's Computer Science seniors voted him Teacher of the Year 5 years in a row. He has also presented numerous technical papers and short courses at a variety of conferences and workshops. Mike is a member of the ACM, SIGGRAPH, IEEE, and ASME. He has served on a number of conference committees, including co-chairing the 1991 SIGGRAPH international conference and the 2001 IEEE Visualization conference.

Future robotic missions are being planned in order to explore permanently dark regions of the moon, located in craters near the poles. There the rover will monitor for the presence of hydrogen concentration, and collect and analyze samples in order to verify the presence of water ice, and, in addition, and determine the spatial distribution of the ice, including location, depth, and concentration. Unlike the case with Mars, short communication delay to the moon makes safeguarded teleoperation of the rover's surface operations a viable strategy for navigation and control in general. Automated planning systems on the ground could also assist scientists in generating waypoint-based exploration routes. Nonetheless, the intermittent loss of direct line-of-sight communication near the poles justifies considering an approach that combines autonomous on-board decision making with teleoperation. This talk describes an approach for autonomously constructing, executing and revising plans for multiple sorties into and out of cold traps on craters. The approach combines mission ground planning with on-board execution and plan revision.

Biography

Robert Morris is researcher in Computer Science in the Exploration Technology Directorate, Intelligent Systems Division at NASA Ames Research Center. He is involved in a number of efforts that involve the application of advanced AI technology in planning, scheduling and plan execution to the next generation of NASA’s exploration systems. He is currently Principal Investigator of a project for coordinating the sensing activities of distributed, remote sensor webs for measuring the Earth's ecosystems. He has published a number of papers on temporal constraint-based reasoning for automated planning and scheduling. For four years (2000-2004) he was sub-program manager in NASA’s Intelligent Systems Program (IS) in the area of automated reasoning, which focused on developing software capabilities in intelligent sensing, planning and scheduling, health management, V&V of intelligent systems, and coordination of distributed autonomous systems. For 12 years prior to joining NASA, he was professor of Computer Science at the Florida Institute of Technology, during which time he attained the rank of full professor.

This talk will share some of our work around many core and mobile computing with more focused attention to the technologies being developed in his lab, including network processing software, optical networking, wireless USB, ultra-wideband, mobile wireless LAN and WAN, broadband in and to the home, robust networks, trust and security, and wireless sensor networks.

Biography

As director and general manager for the Communications Technology Lab, Alan Crouch is responsible for driving Intel's network technology leadership. Lab focus areas include network processing software, optical networking, wireless USB, ultra-wideband, mobile wireless LAN and WAN, broadband in and to the home, robust networks, trust and security, and wireless sensor networks. The lab is strongly committed to standards-based solutions. Members of the lab are active participants in 3GPP, IEEE, IETF, the Network Processing Forum, and many other communications and networking standards-setting bodies. Crouch has led Intel R&D efforts in network architectures since 1999. His teams have made key contributions to the Network Processing Forum and IETF ForCES network processing industry standards efforts, and the Intel® IXA Portability Framework. His team co-developed the Microblock software architecture for Intel® IXP network processors, to reduce the time and effort for developing network processing applications. More recently, Crouch's teams have delivered platform networking innovations that are being incorporated into Intel® I/O Acceleration Technology and Intel® Active Management Technology roadmaps. In addition, Crouch launched Intel's Communications Technology Lab in the Intel China Research Center, Beijing, and led advanced R&D external engagements with key networking fellow travelers Alcatel, Cisco, Ericsson, and Hewlett-Packard. Before joining Intel, Crouch worked for Network Computer Devices, Inc. and Tektronix, Inc. In 12 years at Tektronix, he held positions of increasing responsibility starting as a software engineer and eventually holding the title of Director of Engineering from 1996 to 1998. Crouch graduated with honors from Oregon State University and holds a bachelor of science degree in computer science.

While Moore's law is alive and well in silicon scaling technology, it is clear that microprocessors have encountered significant technical issues that will influence the overall direction of the future architectures. This talk discusses the recent history of Intel microprocessors, some of the rational that guided the development of those processors. Further, the talk highlights why the future microprocessor architectures will likely look different from the past. The traditional microprocessor architecture uses hardware techniques such as out-of-order processing to extract higher performance out of applications that have little or no explicit parallelism. The hardware techniques employed in the past have continued to improve performance, but at the cost of significantly increasing the power consumption of the traditional microprocessors. The power increases have led to not only higher electrical power delivery costs, but higher costs dissipating the power, resulting in high ambient noise, larger enclosure and hotter laps. To avoid a future that requires asbestos based jeans to properly handle laptops, the microprocessor architecture must change to facilitate higher performance without significantly higher power. It is likely that microprocessor architecture will evolve from the ubiquitous single core, single threaded machine that we know and love, to an architecture that employs more cores and more threads. This shift is apparent in today's market where general purpose processors have included techniques such as Hyper-Threading Technology and Multi-Core processors. This talk will speculate on some potential next steps for that technology and some of the potential implications on software development.

Biography

Doug Carmean is the Chief Architect in Intel's Visual Computing Group in Oregon. Doug was one of the key architects, responsible for definition of the Intel Pentium 4 processor. He has been with Intel for 17 years, working on IA-32 processors from the 80486 to the Intel Pentium 4 processor and beyond. Prior to joining Intel, Doug worked at ROSS Technology, Sun Microsystems, Cypress Semiconductor and Lattice Semiconductor. Doug enjoys fast cars and scary, Italian motorcycles.

In recent years, the interest of both academia and industry in the field of wireless communications has shifted from the conventional paradigm of centralized (cellular) licensed networks to decentralized wireless structures, such as ad hoc and sensor networks, and to unlicensed spectrum access, usually referred to as cognitive radio. In this talk, two key issues in these emerging areas are addressed: distributed synchronization (self-organization) of decentralized wireless networks and quality-of-service (stability) guarantee in cognitive radio. In the first part of the talk, the problem of distributed time synchronization in decentralized networks is tackled. A large number of applications in such networks is enabled by, or benefit from, the availability of a common time-scale among the participating nodes, e.g., tracking of moving objects via sensor networks and coordinated medium access control. Achieving and maintaining synchronization in decentralized scenarios poses new challenges in terms of scalability and energy efficiency, and offers new opportunities through the interplay with specific distributed estimation/ detection applications. In this context, an interesting solution is investigated that is based on the exchange of local time information among neighboring nodes at the physical layer (i.e., via transmission of a train of common waveforms that follows the local clock). Available analytical results are reported, along with numerical examples that corroborate the main conclusions and lend evidence to some interesting phenomena, such as "small-world" effects of shadowing on distributed synchronization. In the second part of the talk, unlicensed spectrum access (cognitive radio) is studied by focusing on a simple model with two single-user links, one licensed to use the spectral resource (primary) and one unlicensed (secondary or cognitive). According to the cognitive radio principle, the activity of the secondary link is required not to interfere with the performance of the primary within the limits imposed by a given quality-of-service guarantee. The presented analysis aims at studying the impact on the system performance of: i) traffic dynamics; ii) sensing errors due to fading at the secondary link; iii) power allocation at the secondary transmitter based on long-term measurements; iv) cooperation (relaying) between secondary and primary transmitters. Throughout the talk, open problems are emphasized along with opportunities for research.

Biography

Osvaldo Simeone is currently an adjunct professor and Post-Doc researcher at the New Jersey Institute of Technology (NJIT), Newark, NJ. He received the M.Sc. degree (with honours) and the Ph.D. degree in Information Engineering from Politecnico di Milano, Milan, Italy in 2001 and 2005 respectively. His current research interests lie in the field of information theory and signal processing aspects of wireless systems with emphasis on cooperative communications, ad hoc wireless networks, MIMO systems, cognitive radio and distributed synchronization.

The talk will address the problem of modeling and rendering virtual urban environments. In the first part of this talk a procedural approach to modeling of urban environments is presented. I will discuss the use of shape grammars as a basis for computing architectural design. In the second part of the talk I will discuss algorithms to accelerate the rendering of large urban models. I will present methods for quickly detecting hidden parts of the scene (occlusion culling) and algorithms to simplify large urban models using point- based rendering and geometric simplification.

Biography

Peter Wonka joined the CSE faculty of Arizona State University as Assistant Professor in 2004 after two years as a post-doctorate researcher at the Georgia Institute of Technology. He received his Ph.D. in computer science from the Vienna University of Technology in 2001 and a masters degree in urban planning in 2002. His research interests include various topics in computer graphics, especially real-time rendering and procedural modeling. Peter Wonka is a member of the PRISM lab.

Time is an important consideration in the design of systems that sense the physical world. Usually a measurement must be taken at a certain time, at certain time intervals, or at the same time as another measurement. Traditionally, measurement systems have used synchronous hardware to control measurement times. The emergence of geographically distributed measurement systems and sensor networks and the ubiquity of digital networks---from desktop to planetary scale---make it desirable to do measurement using asynchronous, networked devices. The synchronicity needed for measurement is obtained by synchronizing a real time clock within each device. Such synchronization can be performed using widely available wireless services ( e.g. Global Positioning System) or network protocols (e.g. IEEE 1588 Precision Time Protocol). Measurements can then be time triggered, that is, a measurement happens when a real time clock reaches a pre-determined time. Alternatively, measurements can be time stamped, where the time that each measurement occurs is due to external factors but the measurement time is accurately and precisely recorded. The talk concludes with a description of new algorithms for data analysis of time stamped measurements, for example filtering and spectral analysis.

Biography

Lee Barford is a Master Scientist in the Measurements Laboratory of Agilent Laboratories in Palo Alto, California. Lee has a bachelor's degree in Computer Science from Temple University, and an MS and Ph.D. in Computer Science and Operations Research from Cornell University. After a short stint teaching at Cornell, Lee became a Member of the Technical Staff at Hewlett-Packard Laboratories in 1987. In 1996 he was promoted to Senior Member of the Technical Staff. In 2000, HP spun off its technology businesses (measurement, test, semiconductors, chemical analysis, and biotechnology) into a new company, Agilent Technologies. Lee became part of Agilent Laboratories at that time. In 2003, Lee was promoted to Master Scientist. In addition to diagnosis, at Hewlett-Packard Labs and Agilent Labs Lee has worked in a number of other research areas, including mechanical CAD systems, wavelet signal processing, networked sensors, automated X-ray test of electronics, and behavioral modeling of nonlinear dynamical systems. His current research interest is applying Bayesian methods to improving speed and accuracy of measurement and test.

Many algorithms in signal and image processing aim to extract and analyze information from measured data; a reliable model can assist in this task. In a non parametric setting, no a priori assumptions are made and no model is assigned to describe the data. Recent efforts are directed towards exploring and utilizing data structure. Think, for example, of a high-dimensional feature vector whose features are correlated and exhibit linear dependence. The features lies on a lower-dimensional hyperplane and can be described and processed using a lower-dimensional representation. Manifolds offer the capability to extend these ideas from a linear to a nonlinear setting. In manifold learning, we assume that the data lies on a low-dimensional manifold in a high-dimensional space. The dimension and structure of the manifold can be learned from the data to assist in various learning tasks. Under the sparse signal framework, signals of interest can be represented using a small subset of elements from a large dictionary. While the dictionary can span the entire high-dimensional space, sparse signals reside in a small portion of it. This assumption can be incorporated into many signal processing problems to yield innovative algorithms that can analyze the data more accurately. This talk presents an extension of manifold learning techniques to supervised and semi-supervised learning with applications to medical diagnosis, hyper spectral imaging, and localization in sensor networks. Additionally, novel sparse reconstruction algorithms with application to magnetic resonance force microscopy will be presented.

Biography

Raviv Raich received the B.Sc. and M.Sc. degrees in electrical engineering from Tel-Aviv University, Tel-Aviv, Israel, in 1994 and 1998, respectively and the Ph.D. degree in electrical engineering from Georgia Institute of Technology, Atlanta, Georgia, in 2004. Between 1999 and 2000, he worked as a researcher with the communications team, Industrial Research Ltd., Wellington, New Zealand. Currently, he is a postdoctoral research fellow at the University of Michigan, Ann Arbor, Michigan. His main research interest is in statistical signal processing with specific focus on manifold learning, sparse signal reconstruction, and adaptive sensing. Other research interests lie in the area of statistical signal processing for communications, estimation and detection theory.

A decade ago, the introduction of turbo codes and iterative message passing algorithms revolutionized the theory and practice of coding. In the ensuing years, the coding theory community has become adept at designing codes from good graphical models - that is, models which imply low-complexity, near-optimal iterative message passing algorithms. Specifically, modern codes are constructed by connecting a large number of simple local codes together via a rich, random-like, cyclic interconnection network. A key observation from this work is that the introduction of cycles to graphical models can enable massive complexity reductions in model, and thus decoding, complexity. Whereas constructive graphical modeling problems (e.g. code design) have been widely addressed by the coding theory community, less is understood about the inverse problem of model extraction. Specifically, can good cyclic graphical models be obtained for existing algebraic codes, or more generally, for arbitrary systems? What tradeoffs exist between model complexity and cyclic topology for a given code? If good models can exist, how can they be obtained, or extracted? This talk presents a theoretical framework for the study of extractive graphical modeling problems. We first examine the limits of extraction by providing a characterization of the tradeoff between cyclic topology and complexity in graphical models for linear codes. Inasmuch as the cyclic topology of a graphical model is related to the performance of the decoding algorithms it implies, the bound presented in this talk provides insight into the limits of graphical model extraction. We then provide a formalization of extraction as optimization and describe some novel heuristics for both defining and solving this optimization problem. We conclude with a discussion of the importance of cyclic model extraction outside of coding.

Biography

Thomas R. Halford received the B. A. Sc. degree in engineering physics from Simon Fraser University, Burnaby, B.C., Canada, in 2001. He is currently a doctoral candidate at the University of Southern California, Los Angeles, where his research focuses primarily on graphical models of codes. He spent the summer of 2005 visiting the Natural Language Processing Group at IBM T. J. Watson Research Center, Yorktown Heights, NY.

This talk first examines distributed computing from a languages/software engineering standpoint, and subsequently uses the distributed programming domain to motivate program generation: a general and powerful approach to automating programming tasks. We begin with a simple-to-state but important problem: how to define middleware that allows the programmer to think of a distributed application largely as a centralized one, while maintaining efficiency and expressiveness. Our NRMI middleware facility is the first to support call-by-copy-restore semantics for arbitrary pointer-based data, letting the programmer think of a remote procedure call much like a local one in many useful cases. We show that NRMI is significantly easier to use than traditional RPC middleware, yet maintains efficiency and full programmer control. If we take the task of simplifying distributed programming to the extreme, we encounter systems that allow unsuspecting programs to execute in a distributed environment. Typically such systems use program generation extensively. We briefly present our J-Orchestra system for automatically enabling Java programs to execute in a distributed setting. We then discuss the impact that the J-Orchestra program transformation techniques have had on a large open-source project (JBoss).

Biography

Yannis Smaragdakis got his Ph.D. from the University of Texas at Austin and is currently an Associate Professor at the University of Oregon. His interests are on the systems and languages side of software engineering. More information on his work can be found at: http://www.cs.uoregon.edu/~yannis

Developing intelligent wireless sensor networking techniques enables us to realize the prominent potential of sensor networks in managing our physical environment. Many characteristics unique to sensor networks, however, pose great challenges to achieving this goal. One of them is the inherent data redundancy, specifically, high spatial correlation, among sensor data that demands cognitive retrieval strategies. In this talk, I will focus on the interconnection between data retrieval and inference in a large-scale reach-back sensor network, by considering the problem of reconstructing a signal field with data retrieved using deterministic scheduling or random access. We use an integrated approach of networking and sensing to quantify how different retrieval schemes influence the reconstruction performance. The proposed approach not only establishes the connection of retrieval schemes to data sampling schemes, but also provides a communication and networking perspective of the traditional sampling problem unique to sensor networks. I will present the results in both infinite and finite sensor density conditions. The infinite density results demonstrate the fundamental difference between distinct data retrieval schemes under the ideal condition. The finite density case, on the other hand, reveals the sensitivity of different retrieval schemes in reality where practical imperfections may exist.

Biography

Min Dong received the B.Eng. degree from Tsinghua University, Beijing, China, in 1998, and the Ph.D. degree in Electrical Engineering with minor in Applied Mathematics from Cornell University in 2004. Since 2004, she has been with the Corporate Research and Development, Qualcomm Inc., San Diego, CA, where she has actively contributed to the design and standardization of an orthogonal frequency division multiple access (OFDMA) system for the evolved 3G broadband wireless communications. Dr. Dong received the IEEE Signal Processing Society Best Paper Award for her work on the pilot design for wireless transmission. Her research interests include communications, signal processing, and mobile networks.

A main challenge in multihop wireless networking is to ensure the reliable and efficient transmission of mission-critical data across multiple hops of unstable wireless links. Medium access control schemes based on random contention (such as IEEE 802.11 DCF) improve the channel-usage efficiency over deterministic scheduling schemes (such as TDMA and OFDMA), but they lack stability in quality-of-service provisioning. In this talk, we argue that the rank-based approach to medium access provides the middle-ground and a proper balance between stability and efficiency for multihop wireless networks. We first discuss random rank-based packet scheduling, to maintain both fairness and efficiency in data transmission over multihop wireless contention. We study a medium access control protocol that makes use of granule time slots and sequences of pseudo-random numbers to promote spatial reuse and to distribute the throughput fairly among nodes. The protocol has two versions: Randomly Ranked Mini Slots (RRMS) utilizes control-message handshakes similar to IEEE 802.11; Randomly Ranked Mini Slots with Busy Tone (RRMS-BT) provides higher throughput but requires a receiver busy tone. Using computer simulation, we demonstrate the performance of this protocol based on appropriate metrics of long-term and short-term fairness. With both fixed and random topologies, we show that these results are robust to difficult network configurations and unsynchronized clocks. We then discuss deterministic rank-based packet scheduling in a multihop wireless network with end-to-end delay constraints. The emphasis is to determine the proper relative weights assigned to the remaining distance and the remaining lifetime in order to rank the urgency of a packet. We consider a general class of transmission schemes that represents such relative weights using a single lifetime-distance factor. We present an analytical framework to study the effect of the lifetime-distance factor on packet loss probability in a general multihop environment, with different configurations of peer-node channel contention. We demonstrate quantitatively how the proper balance between distance and lifetime in a transmission schedule can significantly improve the network performance, even under imperfect schedule implementation.

Biography

Ben Liang received honors simultaneous B.Sc. (valedictorian) and M.Sc. degrees in electrical engineering from Polytechnic University in Brooklyn, New York, in 1997 and the Ph.D. degree in electrical engineering with computer science minor from Cornell University in Ithaca, New York, in 2001. In the 2001 - 2002 academic year, he was a visiting lecturer and post-doctoral research associate at Cornell University. He joined the Department of Electrical and Computer Engineering at the University of Toronto as an Assistant Professor in 2002. His current research interests are in mobile networking and multimedia systems. He received an Intel Foundation Graduate Fellowship in 2000 toward the completion of his Ph.D. dissertation, the Best Paper Award at the IFIP Networking conference in 2005, and the Runner-up Best Paper Award at the International Conference on Quality of Service in Heterogeneous Wired/Wireless Networks in 2006. He is a senior member of IEEE and a member of ACM and Tau Beta Pi.

Wireless sensor networks are envisioned to have a lot of application areas, such as military, homeland security, environment, agriculture, heath care, manufacturing, and so on. The primary functionality of a sensor network is to monitor the environment and transmit the data to a base station for further analysis. Thus, routing is an essential operation in sensor networks. For sensor networks deployed in hostile environments (such as military battlefields), security is critical to ensure privacy, integrity, authenticity, and availability of information and communications. Past researches on sensor network routing focused on efficiency and effectiveness of data dissemination. Few of them considered security during the design phase of a routing protocol. Furthermore, previous work on sensor networks mainly considered homogeneous sensor networks, that is, all sensor nodes are modeled to have same capabilities. Several literatures have shown that homogeneous ad hoc networks have poor fundamental performance limits. To achieve better performance and security, I adopt a Heterogeneous Sensor Network (HSN) model. In this talk, I will present a secure routing protocol for HSNs, which is energy efficient and robust to sensor node failures. Our security analysis demonstrates that the secure routing protocol can defend typical attacks on routing. Our simulation results show that the secure routing protocol has better performance than a popular sensor network routing protocol – Directed Diffusion. At the end, I will talk about my research interests and teaching interests.

Biography

James Du is an Assistant Professor in the Department of Computer Science, North Dakota State University. Dr. Du received his B.E. degree from Tsinghua University, China in 1996, and his M.S. and Ph.D. degrees from the University of Maryland, College Park in 2002 and 2003, respectively, all in Electrical Engineering. He was awarded a fellowship from the University of Maryland towards his Ph.D. study. Dr. Du’s current research is supported by the NSF and NASA. His research interests are heterogeneous wireless sensor networks, wireless networks, computer networks, security, systems and controls. Dr. Du has published more than 40 papers in these areas, and one of the papers received the Best Paper Award from the IEEE MASS 2006 conference. Dr. Du is an Associate Editor of Wiley Journal on Wireless Communication and Mobile Computing, and International Journal of Sensor Networks (InderScience). He has served as a Program Chair and Technical Program Committee (TPC) for a number of major IEEE/ACM international conferences.

Devices should be perceptive, and respond directly to their human user and/or environment. In this talk I'll present new computer vision algorithms for fast recognition, indexing, and tracking that make this possible, enabling multimodal interfaces which respond to users' conversational gesture and body language, robots which recognize common object categories, and mobile devices which can search using visual cues of specific objects of interest. As time permits, I'll describe recent advances in real-time human pose tracking for multimodal interfaces, including new methods which exploit fast computation of approximate likelihood with a pose-sensitive image embedding. I'll also present our linear-time approximate correspondence kernel, the Pyramid Match, and its use for image indexing and object recognition, and very recent work on transfer of predictive structures from previous category recognition tasks. I'll show interface examples based on these techniques, including prototypes for grounded multimodal conversation and mobile image-based information retrieval.

Biography

Trevor Darrell leads the Vision Interface Group at the MIT Computer Science and Artificial Intelligence Laboratory. His interests include computer vision, multimodal interfaces, and machine learning. Prior to returning to MIT he worked as a Member of the Research Staff at Interval Research in Palo Alto, CA, researching vision-based interface algorithms. He received his PhD and SM from MIT in 1996 and 1991, respectively, while working at the Media Laboratory, and the BSE from the University of Pennsylvania in 1988, where he worked in the GRASP Robotics Laboratory.

The next generation of networks is expected to support a wide variety of applications in dynamic, interference-constrained, unreliable, and resource-limited network settings. These systems with such diverse demands necessitate a radically new approach to network design urging a cross-disciplinary effort that spans various areas of information sciences, decisions and systems engineering, computer science, mathematics, and economics. This talk will demonstrate the success of this unifying approach through the discussion of our recent works in the area of `Network Coding'. Network coding is a novel data transmission strategy that promises significant throughput gains over traditional routing strategies. Its strength is based on its guarantee of achieving the maximum possible flow rate of the network for multicast sessions. This is in contrast to simple routing strategies which may yield arbitrarily low utilization of network resources. Although there has been much recent research on network coding, significant questions remain on its practical implementation and a fundamental understanding of its performance. In this talk, I will describe our recent works that have provided answers to both of these vital questions. First, I will present a new algorithm for implementing intersession network coding in general networks. The main innovation would be to show that the ideas underpinning network coding can be extended to serve multiple sessions by allowing coding across sessions. I will present a novel routing/scheduling/coding algorithm that provides a practical method to apply network coding to multiple point-to-point sessions. This algorithm is shown to achieve any rate within the largest known achievable rate region for intersession network coding. Second, the challenging problem of the delay performance of network coding will be addressed. I will introduce a key scenario that will be used to reveal the significant delay gains obtained from network coding over traditional strategies. This is the first work that quantifies these gains. Further, I will explain how these gains translate into economic benefits in a dynamic setting. I will conclude with a brief overview of several other projects inspired by the unifying approach that also led to fundamental and surprising findings.

Biography

Atilla Eryilmaz received his M.S. and Ph.D. degrees in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign in 2001 and 2005, respectively. Since 2005, he has been working as a Postdoctoral Associate at the Laboratory for Information and Decision Systems at the Massachusetts Institute of Technology. His research interests include wir