FALL SEMINAR 2006

(Organized Professor Qing Zhao)

September 29 “Scalable Data Gathering In Wireless Sensor Networks” Professor Bhaskar Krishnamachari, Electrical Engineering Systems, University of Southern California Abstract: We will examine three interesting theoretical research problems pertaining to the energy efficient collection of information from very large scale wireless sensor networks. The first pertains to joint routing and compression, where we will see that while the optimum strategy depends dynamically on the level of correlation in the environment, there exists a simple correlation-unaware near-optimal strategy that is in fact asymptotically optimal. In the second part, we will discover some fundamental scaling laws for data-centric storage and querying mechanisms in the form of conditions under which arbitrarily large sensor networks can be sustained with bounded per-node energy requirements. Finally, we will investigate how the performance of random walk-based queries can be drastically improved in heterogenous networks. Bio: Bhaskar Krishnamachari is an Assistant Professor in the Dept. of Electrical Engineering-Systems at USC, where he holds the Philip and Cayley MacDonald Early Career Chair. His research focuses on the analysis and design of efficient data gathering and self-configuration algorithms for wireless sensor networks. He received the 2004 NSF CAREER award and the USC Viterbi School of Engineering's outstanding junior faculty award in 2005. He serves on the editorial boards of the Ad Hoc Networks journal, the ACM Mobile Computing and Communications Review and the EURASIP Journal on Wireless Communications and Networking and as the Sensor Networks Vice Chair for IEEE ICDCS . He is the author of a textbook titled "Networking Wireless Sensors", published by Cambridge University Press. October 6 “Start Something: Some Thoughts on Entrepreneurship” Nick Triantos, Sand Hill Angels, Inc., Silicon Valley Abstract: I will provide an introduction to some of the issues entrepreneurs will face when they start a company. The talk will focus on legal, financial, and product challenges; the engineering issues are left for each of you to solve when you make the leap into building your own company. I will also try to leave plenty of time for Q&A, since this is such a broad topic. I have 15 years of experience in software engineering roles at various startups, and was most recently Chief Software Architect at NVIDIA. Today, I invest in startups as an angel investor with a group called Sand Hill Angels. Bio: Nick is a Board Director and active member at Sand Hill Angels since 2004. He invests primarily in tech companies, especially those with a focus on high-tech medical devices. Nick has over 15 years of experience in software engineering and engineering management. He was most recently at NVIDIA for 8 years, where he served as Chief Software Architect. Nick is an expert in computer graphics and system software architecture, and also has previous work experience in the set-top box market, console game market, and business productivity software market, from his previous work at Navio (now Liberate), 3DO, and Claris (now part of Apple). These days, Nick is most interested in high-tech medical devices, web-based software applications, and knowledge systems, but he also remains interested in consumer electronics, graphics, and electronic entertainment, too. Nick has a B.S. in computer science from the State Univeristy of New York at Buffalo, and he is currently pursuing MBA's through Columbia University in New York, and the Haas Business School at UC Berkeley. October 20 “Information Processing by Assemblies of Molecules: Directed Self-Assembly, Nonlinear Behavior, Array Architectures” Richard A. Kiehl, Department of Electrical and Computer Engineering, University of Minnesota Abstract: This talk deals with research toward information processing by assemblies molecular components arranged in locally coupled arrays. I will discuss our recent work on 1) self-assembly of 2D nanocomponent arrays by in situ hybridization to DNA scaffolding, 2) the electrical behavior of alkanethiol//oligo(phenylene-ethynylene) bilayer molecular junctions, and 3) information processing by nonlinear phase dynamics in locally connected arrays. This work illustrates the type of paradigm shift that could lead to computers with high-level capabilities far beyond those of conventional systems. Bio: Richard A. Kiehl received the B.S., M.S., and Ph. D. degrees from the School of Electrical Engineering, Purdue University. From 1974 to 1980 he was a member of technical staff at Sandia National Laboratories, where he initiated studies on optical control of microwave semiconductor devices. In 1980 he joined AT&T Bell Laboratories, Murray Hill, as a member of technical staff. He was a leading contributor to the Bell Labs research on heterostructure electronics, particularly heterostructure field-effect transistors. In 1985 he joined IBM as research staff member at the T. J. Watson Research Center and focused his work on III-V and SiGe-based heterostructure CMOS circuitry. In 1993 he became assistant director of the Quantum Electron Device Laboratory at Fujitsu Laboratories Ltd. in Atsugi, Japan, where he lead research on nanoelectronics. He was on the faculty of Stanford University as acting professor of Electrical Engineering from 1996 to 1999, and he is currently the Louis J. Schnell Professor of Electrical & Computer Engineering at the University of Minnesota. He served as associate editor of IEEE Electron Device Letters and was co-editor of the book ¡°High Speed Heterostructure Devices¡± of the Academic Press Semiconductor and Semimetals Series. Professor Kiehl is a member of the American Physical Society and the American Chemical Society and a Fellow of the Institute for Electrical and Electronics Engineers. October 27 “System Theoretic Foundations for Sensor Networks” Venugopal V. Veeravalli, Electrical and Computer Engineering, University of Illinois - Urbana-Champaign Abstract: Networks of distributed wireless sensors capable of collecting, storing, and disseminating a variety of environmental data have the potential to enable the next revolution in information technology. Research to date on such sensor networks has largely been focused on techniques for building the sensors, and on self-configuring protocols for establishing communication between them. However, in order to fully exploit their potential, a core system-theoretic framework for the design, analysis and application of sensor networks is needed. Such a system-theoretic framework is being developed by the research community in the context of a primary application area for sensor networks, namely statistical inference --e.g., detection, stimation, tracking, etc. This presentation will describe recent results in this area and outline some open problems and challenges that lie ahead. Bio: V.V. Veeravalli is a Professor in the department of Electrical and Computer Engineering, and a Research Professor in the Coordinated Science Laboratory at the University of Illinois at Urbana-Champaign. Dr. Veeravalli served as a program director for communications research at the U.S. National Science Foundation in Arlington, VA from 2003-2005. His research interests include distributed sensor systems and networks, wireless communications, detection and estimation theory, and information theory. He is a Fellow of the IEEE, and a recipient of the IEEE Browder J. Thompson Best Paper Award and the Presidential Early Career Award for Scientists and Engineers (PECASE). November 17 "Nanoscale Transistors: Speed or Talent?” Professor Savas Kaya, School of EE & CS, Ohio University Abstract: The minimum feature size in today's CMOS-based integrated circuits has already reached to sub-100nm scale. These systems are also the most complicated, nano-scale 'functional' products ever fabricated. The experience gained and problems encountered in such large-scale systems are invaluable in guiding the evolution of nanosystems. An important class of problems in this regard is associated with the fluctuation phenomena, which deal with the stochastic variation of device performance due to microscopic mismatches among the fabricated devices. It is well understood now that random distribution of dopants causes intrinsic fluctuations in device characteristics and can impede the traditional scaling approach in sub-100 nm MOSFETs. However, other stochastic effects that do not scale with geometry, such as oxide thickness variations (OTV) or line edge roughness (LER) can also contribute to intrinsic fluctuations. In fact the granularity of matter is a fundamental limit that will persist even in well-scaled undoped devices immune to fluctuations due to discrete electrical charges. An efficient way to combat fluctuations in nanotransistors is to either include redundancy in the system design or allow corrective actions for system performance. In this talk, I will give examples of tunable analog and reconfigurable digital circuits built using DG-MOSFETs, a new class of devices that allow dynamic control of threshold voltage. I will emphasize that, in light of parameter fluctuations, power limits and area concerns, the future nano-systems must have reconfigurable architectures, which can already be addressed using DG-MOSFETs, without having to wait for fundamental material breakthroughs. Bio: Savas Kaya (M¡¯01) graduated from Istanbul Technical University in 1992 with a BSc in Electronics and Communication Engineering, received M.Phil. degree in 1994 from the University of Cambridge, U.K., and Ph.D. degree in 1998 from Imperial College of Science, Technology & Medicine, London, U.K., for his work on strained Si quantum wells on vicinal substrates. From 1998 to 2001, he was a Postdoctoral Researcher at the University of Glasgow, Scotland, U.K., carrying out research in transport and scaling in Si/SiGe MOSFETs, and fluctuation phenomena in decanano MOSFETs. He is currently with the Russ College of Engineering, Ohio University, Athens, OH. He has served as Air Force Office of Scientific Research Summer Faculty Fellow in . He has over 30 journal papers and 45 conference proceedings. His other interests include nanocircuits, TCAD, transport theory, nanostructures, process integration, ionic transport and biomolecular modelling in trans-membrane proteins. Dr. Kaya was a member of the organizing committee for IWCE¡¯7, 2000, and IEEE Nanotech¡¯6, . December 1 “mm-wave Transistors & Integrated Circuits” Professor Mark Rodwell, Electrical and Computer Engineering, University of California, Santa Barbara Abstract: I will summarize the following activities: InP bipolar transistors and ICs based upon them: Our present best devices, at 250 nm emitter width, exhibit 650 GHz power gain cutoff frequencies. Efforts are under way to scale these to 125 nm and 65 nm, with the goals of 440 GHz digital clock rates (divider metric) and ~700 GHz amplifiers. We are currently designing 340 GHz power amplifiers with 250 nm devices. We are also desiging microwave Op-amps with very low simulated intermodulation distortion obtained by strong negative feedback III-V MOSFETs: We have started a collaboration with 10 other research groups with the goal of developing MOSFETs with InGaAs and InP channels. We will briefly summarize both the potential benefits and limitations of compound semiconductor channels for MOS VLSI. mm-wave SiGe and CMOS ICs Silicon MOSFETs now have bandwidth sufficient for 100 GHz amplification, and SiGe devices substantially more. While reduced cost in high-volume applications is a worthwhile motivation for mm-wave Si, a better motivation would be new and novel applications enabled by the availability of very large numbers of transistors operating at very high frequencies. Several examples will be discussed, including very large monolithic array transmitters, mm-wave sensor networks employing radar/imaging techniques, and mm-wave communications links employing imaging techniques for spatial-division multiplexing (mm-wave-MIMO). Bio: Mark Rodwell is Professor in the Electrical and Computer Engineering Department at UCSB. He received his Ph.D. in Electrical Engineering from Stanford University in 1988. He worked at AT&T Bell Laboratories during 1982-1984. His research group works to extend the operation of electronics to the highest feasible frequencies. Particular interests include InP bipolar transistors and high frequency IC design in both III-V and Silicon VLSI. His group's work on GaAs Schottky-diode ICs for subpicosecond / mm-wave instrumentation was awarded the 1997 IEEE Microwave Prize. Prof. Rodwell was elected IEEE Fellow in 2003.