Fall 2010 Seminars

Demystifying DoD Research Funding

Dev Palmer
US Army Research Office
Wednesday, December 8, 2010, 4:00-5:00PM, Marston 132
Host: Do-Hoon Kwon

From the outside, the DoD science and technology funding infrastructure can appear complex and obscure. Without guidance it is difficult to answer some basic questions such as: Who decides which research topics are important to the DoD? What process determines how DoD research funds are allocated? How can I access this information? And most importantly, how can I maximize the likelihood of obtaining funding for my research? Formulated from experience both as a contract researcher and as a government program manager, this talk will cover some of the information, ideas, disciplines, and techniques that can help demystify DoD research funding.

Dr. Dev Palmer is a Program Manager at the US Army Research Office in
Research Triangle Park NC, responsible for extramural basic research in
electromagnetics, microwaves, and power. From 1991 to 2001, he served on the technical staff at the MCNC Research and Development Institute. He is
engaged in antenna research as a Member of the Graduate Faculty in the Duke University Department of Electrical and Computer Engineering and
occasionally teaches introductory electromagnetics. He received the Ph.D.
degree in Electrical Engineering from Duke in 1991, is a registered Professional Engineer, and an IEEE Senior Member.


Is Moore’s Law Going to Die before Himself? ---- A Personal Perspective
Qiangfei Xia
UMass Amherst/ECE Dept.
November 22, 4:00pm, Marston 132
Host: Do-Hoon Kwon

While Gordon Moore is enjoying fishing in Hawaii, the rest of the world is not on vacation. Following an empirical rule that bears Moore’s name, Intel just announced that it is investing billions in 22-nanometer chip production. ‘Moore’s Law’, which predicts that the number of transistors on a chip will double every two years, has become a principle for the IC industry in delivering more and more powerful semiconductor chips at decreasing cost per transistor.

However, there is plenty of but not unlimited room at the bottom. Continuing evolution of electronics beyond the scaling limit requires revolutionary vision and broad thinking across disciplines. What novel materials are going to replace silicon? What enabling technologies will bring us breakthroughs? Are there any simpler but more capable devices? Do we really have to design our computer in a planar fashion? To give ‘Moore’s Law’ a boost, new paradigms are being developed in all areas. And part of the efforts will be covered in this seminar.

Dr. Qiangfei Xia is an assistant professor of Electrical and Computer Engineering at UMass Amherst. He received his Ph.D. in Electrical Engineering from Princeton University in 2007.  He then worked at HP Labs in Palo Alto, CA for 3 years before moving back to the east coast. His research interests include advanced nanofabrication/integration technologies and novel nanodevices for the post-CMOS nanoelectronic systems.


Overview of Radio Astronomical Receivers
Gopal Narayanan
UMass Amherst/Astronomy department
Friday, November 19, 2010, 11:15am - 12:05pm, Gunness Student Center Conf. Room
Host: Joe Bardin

I will review the current state of art in radio astronomical receivers. At radio, millimeter and submillimeter wavelengths, observations with radio telescopes yield a very different picture of the universe compared to optical wavelengths. Radio Astronomy is poised in an exciting era of very large telescopes and telescope arrays. To make full use of the investment in such large telescope projects, sensitive receivers are needed. Radio Astronomical Receivers are at the confluence of advances in RF, microwave, millimeter, superconducting, digital signal processing technologies, and advances in computational power to harness these technologies. In other words, working in radio astronomical receivers is a great example of applied EE! In the talk, I will review some science, and then talk about continuum and spectroscopic receivers that are being built to address the science.

Gopal Narayanan is a Research Associate Professor in Astronomy at the University of Massachusetts, Amherst. He has a Bachelor's degree in EE from Anna University, India, and a Master's Degree in EE from Caltech. His PhD in Astronomy is from the University of Arizona, Tucson. After doing his postdoctoral research at the Five College Radio Astronomy Observatory at UMass, he joined their faculty ranks in 2000, where he has been ever since. Prof Narayanan's research interests range from millimeter/submillimeter receivers for radio astronomy to the study of star formation in molecular clouds in our own galaxy and in other galaxies. He is currently working with other collaborators in the Astronomy Department at UMass to commission the Large Millimeter Telescope (LMT), a 50-meter diameter radio telescope that is being built in Mexico.


Progressive Inter-carrier Interference Equalization for OFDM Transmission over Time-varying Underwater Acoustic Channels
Shengli Zhou
Univ. of CT
Friday, Nov. 5, 2010, 11:15am, Gunness Student Center
Host: Hossein Pishro-Nik

Multi-carrier modulation in the form of orthogonal-frequency-division-multiplexing (OFDM) has been intensively pursued for underwater acoustic (UWA) communications recently due to its ability to handle long dispersive channels. Fast variation of UWA channels destroys the orthogonality of the subcarriers and leads to inter-carrier interference (ICI), which degrades the system performance significantly.

In this talk, I will present a progressive receiver dealing with time-varying UWA channels. The progressive receiver is in nature an iterative receiver, based on the turbo principle. However, it distinguishes itself from existing iterative receivers in that the system model for channel estimation and data detection is itself continually updated during the iterations. When the decoding in the current iteration is not successful, the receiver increases the span of the ICI in the system model and utilizes the currently available soft information from the decoder to assist the next iteration which deals with a channel with larger Doppler spread. Numerical simulation and experimental data collected from the SPACE08 experiment show that the proposed receiver can self adapt to channel variations, enjoying low complexity in good channel conditions while maintaining excellent performance in tough channel conditions.

At the end, I will describe our recent progress towards a practical underwater acoustic modem.

Shengli Zhou received the B.S. degree in 1995 and the M.Sc. degree in 1998, from the University of Science and Technology of China (USTC), Hefei, both in electrical engineering and information science. He received his Ph.D. degree in electrical engineering from the University of Minnesota (UMN), Minneapolis, in 2002. He is now an associate professor with the Department of Electrical and Computer Engineering at the University of Connecticut (UCONN), Storrs. He holds a United Technologies Corporation (UTC) Professorship in Engineering Innovation.

His general research interests lie in the areas of wireless communications and signal processing. His recent focus is on underwater acoustic communications and networking. Dr. Zhou has served as Associate Editors for IEEE Transactions on Wireless Communications, Feb. 2005 - Jan. 2007, and IEEE Transactions on Signal Processing, Oct. 2008 - Sept.  2010, and is now an associate editor for IEEE Journal of Oceanic Engineering. He received the 2007 ONR Young Investigator award and the 2007 Presidential Early Career Award for Scientists and Engineers (PECASE).


Hardware Evaluation of Functions using Optimized Polynomial Approximations
Arnaud Tisserand
Wednesday., Nov. 3, 2010, 4:00 pm, Marston 132
Host: Maciej Ciesielski

The design of high-performance arithmetic operators is an important issue in application specific integrated circuits (ASICs), systems on chip(SoCs) and field-programmable gate arrays (FPGAs) implementations.  Basic operations such addition/subtraction and multiplication have always been implemented as high-performance operators in digital circuits. Many recent applications require fast evaluation of more complex operations such as division, reciprocal, square-root, inverse square-root, trigonometric functions (sine, cosine, arc-tangent, hyperbolic functions...), logarithm or exponential. Evaluation of compositions of such functions is also required in some applications. A wide variety of algorithms for the approximation of functions has been proposed.  Among them table based methods, polynomial or rational approximations, digit-recurrence algorithms and combinations of these solutions are the most commonly used.

Function approximation is often performed using polynomial evaluation in software as well as in hardware implementations. In this presentation, a method that produces polynomial approximations "well-suited" for high-performance hardware implementations is described.  It faces with two problems: the generation of the polynomial coefficients that ensure low approximation errors and the sizing of intermediate computations that provide low round-off errors. The complete method has been implemented as an automatic code generator.  The generated operators are small and fast. They are also numerically validated (with a formal proof) at compile time.

Arnaud Tisserand is a senior researcher in CNRS (French National Center for Scientific Research) in IRISA laboratory in Lannion, France.
Previously, he was a member and former head of ARITH group of the LIRMM (Montpellier Laboratory of Computer Science, Robotics, and
Microelectronics) in Montpellier, France. He was a researcher at INRIA (French National Institute for Research in Computer Science and Control) in LIP Laboratory in Lyon, France. He was a research expert in the Ultra-Low Power Group of the CSEM (Swiss Center for Electronics and Microtechnology) in Neuchâtel, Switzerland. His research interests include computer arithmetic, computer architecture, VLSI and FPGA design, design automation, low-power design and applications in scientific computing, digital signal processing and cryptography.


Challenges and Opportunities for Energy Reduction in Wireless Sensor Networks
Olivier Sentieys
University of Rennes
Monday, Nov. 1, 2010, 4:00pm, Marston 132
Host: Maciej Ciesielski

An energy-efficient platform for wireless sensor networks and its dedicated software framework developed in the Cairn team (University of Rennes - IRISA) are described and optimizations are proposed at both hardware and software level. The hardware is  based on low cost and low power components linked in a multilayer architecture. Some processing, such as error correcting codes, are accelerated by a low power FPGA and the platform offers the possibility to dynamically adapt the voltage and the frequency of the processor. Finally the software system implementing the protocols stack has a very small memory foot print. Some parameters  optimizations are then proposed at lower protocol layers to decrease the global energy consumption of the platform. The medium access protocol relies on an asynchronous rendez-vous scheme initiated by the receiver, and the wake-up interval can be tuned to increase the lifetime of the network. The influence of error correcting codes on the energy consumption is evaluated and some promising cooperative strategies, based on relay and distributed space-time codes are proposed.

"A Complete Design-Flow for the Generation of Ultra Low-Power WSN Node Architectures Based on Micro-Tasking"

Wireless Sensor Networks (WSN) are a new and very challenging research field for embedded system design automation, as their design must enforce stringent constraints in terms of power and cost. WSN node devices have until now been designed using off-the-shelf low-power microcontroller units (MCUs), even if their power dissipation is still an issue and hinders the wide-spreading of this new technology. In this talk, we propose a new architectural model for WSN nodes (and its complete design-flow from C downto synthe- sizable VHDL) based on the notion of micro-tasks. Our approach combines hardware specialization and power-gating so as to provide an ultra low-power solution for WSN node design. Our first estimates show that power savings by one to two orders of magnitude are possible w.r.t. MCU-based implementations.

Olivier Sentieys is a Professor at the University of Rennes, France. He is leading the CAIRN Research Team at INRIA Institute (which is the French national institute for research in computer science and control) and IRISA Laboratory. His research interests include mobile communication systems, finite arithmetic effects, low-power and reconfigurable architectures.

Emerging Technologies for X-ray Imaging of the Breast
Stephen Glick
UMass Medical School
Monday, October 25, 2010, 4:00pm, Marston 132
Host: Sigfrid Yngvesson

Although x-ray mammography has saved many lives and is considered the imaging modality of choice for early detection of breast cancer, it is far from perfect. One of the limiting problems with the conventional method for mammography is that the recorded image represents the superposition of a three-dimensional (3D) object onto a two-dimensional (2D) plane. One approach for improving  the early detection and diagnosis of breast cancer is through tomography which provides a 3D imaging capability. This talk will review new tomographic imaging technologies that are being studied at the University of Massachusetts Medical School.

Dr. Stephen Glick is the director of the Tomographic Breast Imagine Research Lab at the University of Massachusetts Medical School, which focuses on novel x-ray breast imaging techniques including digital breast tomosynthesis and breast CT. His laboratory has published on several aspects of breast CT including radiation dose, imaging technique optimization, and on detection studies for lesions and microcalcifications with breast CT. In particular, this group has fabricated a prototype, dedicated breast CT scanner, and used this scanner to image over 60 mastectomy breast specimens.

Network Science for Wireless Communication: Information Dissemination, Mobility, and Resilience
Edmund Yeh
Yale University
Monday, Oct. 18, 2010, 4:00pm, Marston 132

Over the past decade, there has been a concerted effort to develop a network science for studying physical, biological, social, and information networks within a common framework.  Of particular interest is the understanding of connectivity, information dynamics, and robustness in large-scale networks with spatial location and mobility.  In this talk, we discuss a number of recent results from the application of network science ideas to mobile wireless communication.

We first study connectivity and information dissemination in large-scale wireless networks modelled by random geometric graphs with dynamic on-off links.  Using a percolation-based perspective, we show that the delay for information dissemination exhibits two behavioral regimes, corresponding to a phase transition of the underlying network connectivity. When the dynamic network is in the subcritical phase, ignoring propagation delays, the dissemination delay scales linearly with the Euclidean distance between the sender and the receiver. When the dynamic network is in the supercritical phase, the delay scales sublinearly with the distance.  More interestingly, by using a new analysis which maps a network of mobile nodes to a network of stationary nodes with dynamic links, we show that the above results can be used to characterize information dissemination in wireless networks with mobile nodes.

Next, we study the resilience of wireless networks to node and link failures.  In sensor networks exposed to natural hazards and battery constraints, and in military networks exposed to enemy attack, node and link failures are a common occurrence, and are often correlated. Furthermore, in electrical power networks or wireless computing networks, cascading failure from power blackouts or virus epidemics may result from a small number of initial node failures triggering a correlated failure sequence affecting the whole network.  From the percolation perspective, the resilience of the network can be characterized in terms of whether correlated node failures lead to a large connected component of failed nodes or not.  Using this approach, we obtain analytic conditions on the existence or non-existence of correlated and cascading failures.

Joint work with Zhenning Kong.

Edmund Yeh received his B.S. in Electrical Engineering with Distinction from Stanford University in 1994, his M.Phil in Engineering from the University of Cambridge in 1995, and his Ph.D. in Electrical Engineering and Computer Science from MIT in 2001.  Since 2001, he has been on the faculty at Yale University, where he is currently an Associate Professor of Electrical Engineering, Computer Science, and Statistics.  Professor Yeh is the recipient of a Humboldt Research Fellowship, an Army Research Office Young Investigator Award, the National Science Foundation and Office of Naval Research Graduate Fellowships, the Winston Churchill and Barry Goldwater Scholarships. He is a member of Phi Beta Kappa and Tau Beta Pi.