ECE Seminar Series: Fall 2009

November 20, 2009
Digitally Assisted Architectures for RF Transceivers
Joel Dawson
Massachusetts Institute of Technology
Friday, Nov. 20, 2009, 11am, Gunness Conference Room
Host: Christopher Salthouse (Electrophysics) (Communications and Signal Processing)

This presentation focuses on the principles of good digitally assisted design, and examines a few recent successful architectures. Chief among the examples treated is a thorough analysis and characterization of a new power amplifier linearization architecture, with an exploration of its advantages and limitations.. This architecture represents a new type of hybrid design for linearizers, in which functionality is optimally partitioned between the analog and digital domains. Also discussed are a new technique for extremely low-offset multipliers, mixers, and VGAs, and a new architecture for medical implantable transceivers.

JOEL L. DAWSON is an associate professor in the Department of Electrical Engineering and Computer Science at MIT. He received the S.B. in EE from MIT in 1996, and the MEng. degree from MIT in EECS in 1997.  He went on to pursue further graduate studies at Stanford University, where he received his Ph.D. in Electrical Engineering for his work on power amplifier linearization techniques. Before joining the faculty at MIT, Dr. Dawson spent one year at a startup company that he co-founded. He continues to be active in the industry as both a technical and legal consultant. Prof. Dawson received the NSF CAREER award in 2008, and the Presidential Early Career Award for Scientists and Engineers (PECASE) in 2009.

Members of the Dawson group at MIT pursue solutions to a wide variety of problems in analog, mixed-signal, and RF circuit design.  Our current focus is on RF transceiver architectures for deep-submicron CMOS.  In addition, we work on biomedical device development in collaboration with clinicians at the Beth Isreal Deaconess Medical Center in Boston, MA.

November 19, 2009
College of Engineering Gupta Professorship lecture
How To Implant a Fluorescence Microscope
Christopher Salthouse
Univ. of MA, ECE Dept.
Thursday, Nov. 19, 2009, 4-5pm, Massachusetts Room, Mullin's Center
(CSE) (Communications and Signal Processing) (Electrophysics)

Prof. Salthouse will explain why fluorescence microscopy plays such an important role in biomedical role and how an implantable fluorescence microscope could expand fluorescence microscopy applications even further. Then, he will explain how the integrated circuits which have revolutionized computation will solve this biomedical problem.

Christopher Salthouse came to Massachusetts in 1996 to start his undergraduate studies at the Massachusetts Institute of Technology. In 2000, he received his Bachelor's and Master's degrees in Electrical Engineering from MIT. He spent six more years at MIT working towards his Ph.D. During that time, he developed integrated circuits for cochlear implants, devices that give hearing to the deaf. After finishing his Ph.D., Prof. Salthouse moved to Massachusetts General Hospital where he worked as a research fellow for three years developing novel fluorescence imaging instruments. He came to the University of Massachusetts in August of 2009 where he has founded the Biomedical Electronics Laboratory.

November 13, 2009
Anonymity via networks of mixes
Venkat Anantharam
EECS Department, Univ of CA, Berkeley
Friday, Nov. 13, 2009, 11:15am, Gunness Conference Room
Host:Hossein Pishro-Nik  (Communications and signal processing)
 

Mixes are relay nodes that accept packets arriving from multiple sources and release them after variable delays to prevent an eavesdropper from associating outgoing packets to their sources. We assume that each mix has a hard latency constraint. Using an entropy-based measure to quantify anonymity, we analyze the anonymity provided by networks of such latency-constrained mixes, focusing on the single destination case. Our results are of most interest under light traffic conditions. A general upper bound is presented that bounds the anonymity of a single-destination mix network in terms of a linear combination of the anonymity of two-stage networks. By using a specific mixing strategy, a lower bound is provided on the light traffic derivative of the anonymity of single-destination mix networks. The light traffic derivative of the upper bound coincides with the lower bound for the case of mix-cascades (linear single-destination mix networks).

Venkat Anantharam received the B.Tech in Electronics in 1980 from the Indian Institute of Technology, Madras (IIT-M) and the M.A. and C.Phil degrees in Mathematics and the M.S. and Ph.D. degrees in Electrical Engineering in 1983, 1984, 1982 and 1986 respectively, from the University of California at Berkeley (UCB). From 1986 to 1994 he was on the faculty of the School of EE at Cornell University. From 1994 he has been on the faculty of the EECS department at UCB.

Anantharam received the Philips India Medal and the President of India Gold Medal from IIT-M in 1980, and an NSF Presidential Young Investigator award (1988 -1993). He a co-recipient of the 1998 Prize Paper award of the IEEE Information Theory Society (with S. Verdú) and a co-recipient of the 2000 Stephen O. Rice Prize Paper award of the IEEE Communications Theory Society (with N. Mckeown and J. Walrand). He received the Distinguished Alumnus Award from IIT-M in 2008. He is a Fellow of the IEEE.
 

November 9, 2009
Secured Arithmetic Operators for Cryptography
Arnaud Tisserand
CNRS, IRISA CAIRN, Lannion, France
Monday, November 9, 2009, 4-5pm, ELab 303
Host: Maciej Ciesielski (CSE)

A cryptosystem can be considered in theory hard to break, but in practice the physical implementation of the algorithm may provide weaknesses. For example a hardware implementation of a cryptosystem on a smart card or a FPGA, when the algorithm is executed, can provide "side channel" information (power consumption traces, electromagnetic emissions...) which can help an attacker.

Arithmetic operators are key elements of a crypto-processor. A lot of additions, multiplications, divisions, inversions and exponentiations on very large numbers have to be computed. For instance, elliptic curve cryptography (ECC) requires 160-600 bits numbers on finite fields GF(2^m) or GF(p). The design of efficient arithmetic operators requires very fast algorithms, clever representations of numbers and very careful implementations (FPGA, ASIC, smart cards).  Speed, circuit area and power consumption are not the only parameters for the design of secured arithmetic operators, robustness against side channel and/or fault injection attacks is now another important parameter.

In this talk, we will first introduce the cryptographic context, side channels and fault injection attacks. Then we will present standard methods for the design arithmetic operators. In the last part, we will present solutions for the design of secured arithmetic operators against side channel and/or fault injection attacks.

Arnaud Tisserand is a senior researcher in CNRS (French National Center for Scientific Research) in IRISA laboratory in Lannion, France. Previously, he was a former 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.

Nov. 6, 2009
Physical Image Processing: Efficient Signal Processing in the Retina and in Silicon
Alyosha Molnar
Cornell University
Friday, Nov. 6, 2009, 11am, Gunness Conference Room
Host: Christopher Salthouse

Efficient image processing often involves performing a preliminary set of computations local to the pixels sensing the image itself. In the mammalian retina, scene information is robustly processed, in an analog format before being discretized into spikes for transmission to the brain. This processing, which converts simple intensity information into a more complex, compressed format will be discussed, as well as some of the neural circuits that underlie it.  Strikingly, the retina employs circuit architectures similar to those used in analog integrated circuits to suppress noise and distortion introduced by the components  (synapses) from which those circuits are constructed. Similar levels of complexity and diversity of information can be captured using a new class of pixel, manufactured in standard CMOS, which performs simple, local computations optically before transduction into electrical signals.  Arrays of these pixels have been shown to be  useful in extracting 3-D information from the light field, as well as for performing transformations similar to image processing performed by the mammalian early visual system.

Alyosha Molnar received his bachelors degree in engineering from Swarthmore college in 1997.  After a stint as a deckhand on a fishing boat, he joined Conexant Systems in 1998 as an RFIC design engineer, where he developed the first GSM direct conversion receiver to be sold successfully on the open market.  In 2001 he entered graduate school at UC Berkeley, where he developed the first generation of low power radio transceivers for “Smart dust” and then changed directions and joined a retinal neuroscience lab where he studied the neural circuitry of the mammalian inner retina.  In 2007 he joined the faculty ofCornell  University’s school of electrical and computer engineering.  His present research interests included circuits for neural interfaces, CMOS radio architectures, and novel imaging techniques using structures built in standard CMOS.

November 2, 2009
ViSE:  Incorporating Sensor Networks into GENI's Infrastructure
David Irwin
University of Massachusetts Amherst, Computer Science Dept.
Monday, Nov. 2, 2009, 4-5pm, ELab 303
Host: Michael Zink

In this talk, I will give a broad overview of the NSF GENI initiative---an experimental suite of infrastructure designed to support network science and engineering experiments---and the current state of GENI prototyping.  I will then discuss the ViSE (Virtualized Sensing Environment) testbed we are building as part of the GENI effort, which treats sensor networks as a shared infrastructure for experimentation.  Finally, I will summarize two recent ViSE-related projects:  SRCP and VSense.   SRCP is a narrow control plane protocol for perpetual sensor infrastructure that provides the node-level visibility, accessibility, and interactivity necessary for remote management.  VSense extends the (hardware) virtualization paradigm to share sensors between concurrent applications at the level of individual actuations, while providing each application the abstraction of a dedicated sensor node.

David Irwin is a Postdoctoral Research Associate in the Computer Science Department at the University of Massachusetts, Amherst.  His research focuses broadly on computer systems, including operating systems, wide-area distributed systems, and sensor networks.  He received a MS and PhD in Computer Science from Duke University in 2005 and 2007, respectively, where his research focused on the design of an extensible "operating system" for networked collections of hardware components spread across multiple sites.  This work now serves as a candidate control framework for the NSF GENI initiative.  His current focus is on lowering the barrier to managing sensor networks as a shared infrastructure for experimentation, also as part of the GENI initiative.

October 28, 2009
Impact of Increased Spatio-Temporal Radar Data Resolution on Forecaster Wind Assessments, Warning, and Confidence
Ellen Bass
University of VA
Wednesday, October 28 2009, 4-5pm, ELab 303
Host: Michael Zink

Technological advances could enhance the forecaster warning process by providing higher resolution radar data that sense closer to the ground, have greater spatial resolution, and have faster update rates.  Because wind speed plays a critical role in severe thunderstorm warnings, this talk presents a study investigating the impact of increased spatio-temporal resolution weather radar data on forecaster accuracy of predictions of ground level wind gusts.  It also assesses the impact of these radar data on forecaster confidence and on warning decisions.  The increased resolution radar data are from four CASA radars, an experimental network of X-Band radars in Oklahoma.  In a static case review setting, 30 forecasters with National Weather Service experience evaluated six severe weather cases under two conditions:
a)  using conventional weather radar data and
b)  using both conventional and additional data from the CASA experimental radar network.
Forecasters’ two to five minute predictions of ground level wind gusts were compared to measurements from ground-based wind sensors.  When given the additional radar data, the participants forecasted 20% greater wind speeds, improved the accuracy of their wind assessments by 30%, indicated they were more confident in their assessments, and changed the number of affirmative decisions to warn from 15 to 35.  Implications for the adoption of new technology on warning operations are discussed.

Ellen Bass is an associate professor in the Department of Systems and Information Engineering in the School of Engineering and Applied Science at the University of Virginia.  Her research focuses on understanding and modeling how human operators perform in real-time complex systems in order to inform the systems engineering process:  operational concept definition, requirements for decision support and human-computer interaction, procedures the operators will follow, and training requirements.  She develops analytical frameworks, measures, and methods that quantify total system performance including end users, their tools, features of the task environment and organizational factors.  For over twenty five years, she has been involved in systems engineering research and design with relevant experience in cognitive modeling, cognitive systems engineering, human factors, intelligent decision support and training environments, and simulation.  Bass currently serves as the Vice President of Human-Machines Systems for the IEEE Systems, Man, and Cybernetics Society (SMCS).  She is an associate editor for the journal IEEE Transactions on Systems, Man, and Cybernetics, Part A.  She is the Program Chair Elect of the Cognitive Engineering and Decision Making (CEDM) Technical Group of the Human Factors and Ergonomics Society and a member of the editorial board for the journal Human Factors.  She is a contributing editor of The International Journal of Applied Aviation Studies.

October 21, 2009
Design of High Speed SerDes
Kuan Zhou
University of New Hampshire
Wednesday, October 21, 2009, 4-5pm, ELAB 303
Host: Russ Tessier (CSE)

With the rapid growth of the internet and the development of new storage techniques, larger and larger transmission capacity is required. High speed data transmission is thus necessary. Meanwhile, contrast to the higher density of on-chip transistors, the number of on-chip pads decreases as the chip area is smaller given the same number of transistors. Therefore, the serializer/deserializer (SerDes) is needed here to significantly decrease the number of transmission lines or pins. More importantly, the SerDes chip can eliminate the skew in parallel data by converting it to serial data, which makes the circuit more reliable.

In this presentation we will demonstrate several techniques to improve the SerDes performance, which includes the symmetrical multiplexers, deep trench isolation techniques etc. The successful SerDes research is anticipated to significantly benefit high speed data communication systems that require very high throughput, such as those used in terahertz imaging and chip-to-chip communications.

Dr. Kuan Zhou received his bachelor’s degree from Huazhong Univ. of  Science and Technology in China in 1996, and Master of Science from Chinese Academy of Sciences in 1999. In 2004, he received Master of Science in Computer Engineering and Ph.D. in Electrical Engineering from Rensselaer Polytechnic Institute. Now he is  affiliated with the University of New Hampshire as an assistant professor. Dr. Zhou’s expertise is in integrated circuit (IC) design. He has extensive experience in high-speed FPGA design, self-timed circuit applications, SerDes design and ADCs. His current research interest focuses on the development of low-power aVLSI circuits for neuromorphic engineering systems.

October 19, 2009
Effects of Multi-User Diversity and Correlation on Secrecy Capacity over Fading Channel
Jeongseok Ha
Korea Advanced Institute of Science and Technology (KAIST)
Monday, October 19, 2009, 4-5pm, ELAB 303
Host: Hossein Pishro-Nik (Communications and signal processing)

A. D. Wyner introduced wiretap channel codes to keep legitimate communications secret from the eavesdropper without sharing a secret key. He showed that the perfect secrecy can be achieved by the channel codes if the legitimate receiver (Bob) has a better channel from the transmitter (Alice) than that of the eavesdropper (Eve), which is so called the degraded wiretap channel. From this work, the physical layer security has been under active consideration in recent years to strengthen the existing cryptographical protocols or provide an alternative.

In this talk, we focus on analyzing the maximum rate (i.e., secrecy capacity) to communicate between Alice and Bob reliably while Eve is totally ignorant of the transmitted messages under the wiretap channel framework. We will address such a topic in two different scenarios.

We first consider the correlated wiretap channel where the main (Alice-Bob) and eavesdropper (Alice-Eve) channels are correlated. Assuming that the transmitter knows the full Channel State Information (CSI) (i.e., the channel gains from the transmitter to the legitimate receiver and the eavesdropper), we quantify the loss of the secrecy capacity due to the correlation and investigate the asymptotic behavior of the secrecy capacity at high Signal-to-Noise Ratio (SNR) regime. We have found that the secrecy capacity converges to an upper-bound which is derived in a closed form. Our work clearly shows how the security capacity depends on two channel parameters; the correlation coefficient and the ratio of the main to the eavesdropper channel gains.

Next, we consider the wiretap channel for the forward link of cellular systems in which there are one transmitter and multiple receivers/users who report their channel state information (CSI) to the transmitter. We investigate the impact of the multiuser diversity on the secure communication (i.e., secrecy capacity). To get useful insights, we consider a fixed transmit power case and have some of important performance measures such as outage probability, average secrecy rate, failure rate, and average wait time in closed forms. The analysis interestingly shows that the multiuser diversity becomes always disadvantageous for the secure communications.

Jeongseok Ha (jsha@ee.kaist.ac.kr, +82-42-350-6167) received Ph.D. degree in Electrical and Computer Engineering from Georgia Institute of Technology, Atlanta, in 2003. From 1994 to 1999, he was a researcher with Electronics and Telecommunications Research Institute (ETRI), Daejeon, Korea, where he was involved in developing base stations of IS-95 cellular system and wireless local loop. Currently, he is with Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea as an assistant professor. His research interests include error-control systems, physical layer security, and distributed source coding.

Oct. 16, 2009
Physical Layer Secrecy Encoding Under Eavesdropping Attacks
Shuangquing Wei
Louisiana State University, Baton Rouge
Friday, October 16, 2009, 11:15am, Gunness Conference Room
Host: Dennis Goeckel (Communications and Signal Processing)

Wireless communications is fundamentally more vulnerable to attacks of jamming and eavesdropping than its wireline counterpart, due to the nature of free-air electromagnetic propagation.  With the ubiquitous deployment of wireless devices, study and development of counter-jamming and counter-eavesdropping measures have become more imminent than ever. Unlike the strategies developed in higher layers in wireless networks to cope with malicious attacks, counter-measure mechanisms originated from physical layer enables us to confront with interference and information leakage  more directly and effectively.

In this talk, we will present some of our recent results on counter-measures against intelligent jammers and eavesdroppers.  In the first part of the talk, we will put emphasis on collaboration based Wyner type encoding to achieve perfect secrecy between two legitimate parties over an eavesdropped communication link. We will show how to exploit channel randomness inherent in the feedback channels to attain a strictly positive secrecy rate even when the eavesdropper's channel is less noisy than the legitimate receiver's channel.  In the second part of the talk, we will introduce jamming into the picture where an adversary node is an active eavesdropper.  The active eavesdropper can choose between two functional modes: eavesdropping the transmission between the legitimate parties (Ex mode), and jamming it (Jx mode) -- the active eavesdropper cannot function in full duplex mode. We show that even in the best-case scenario when the transmitter knows the eavesdropper's strategy in advance, such an active eavesdropper can still induce moderate to severe degradation of the achievable secrecy rate.

Dr. Shuangqing Wei obtained his B.E. and M.S. both in Electrical Engineering from Tsinghua University, China, and then his Ph.D. in 2003 at the University of Massachusetts, Amherst.  He is an Associate Professor in the Department of  Electrical and Computer Engineering of Louisiana State University, Baton Rouge.  His  research interests are in the areas of information theory and communication theory, in particular their applications to wireless communication systems and networks such as physical layer secrecy, cognitive radio networks. Dr. Wei is currently serving as an Editor for IEEE Transactions on Wireless Communications and an Associated Editor for IEEE Transactions on Vehicular Technology. He has served as TPC members for numerous major IEEE communication conferences, including ICC, MILCOM, GLOBECOM and WCNC

October 9, 2009
Ballistic Electronics - Breaking the Barrier in Terahertz Speed Processing
Martin Margala
UMass, Lowell
Friday, October 9, 2009, 11:15am, Gunness Conference Room
Host: Sandip Kundu (Electrophysics) and (CSE)

This presentation will describe a major research effort at Dr. Margala's group in developing ultra high frequency switching electronics for terahertz speed computation. The talk will highlight achievements accomplished during the past four years in defining new switching transistor elements, new logic structures to new sequential elements that will make this very high speed computation possible. The talk will conclude with the outlook into the next five years in this newly established field of ballistic nanoelectronics.

Martin Margala (IEEE-SM’04) received the M.S. degree in microelectronics from Slovak Technical University, Slovakia, in 1990 and the Ph.D. degree in electrical and computer engineering from the University of Alberta, Canada, in 1998. He is currently an Associate Professor with the Electrical and Computer Engineering Department, University of Massachusetts, Lowell. Previously, he was with the University of Rochester, Rochester, NY, and with the University of Alberta. From 1998 to 2003, he has been an adjunct scientist with the Telecommunications Research Labs, Edmonton, Canada. He holds three patents (two others pending) and is an author or coauthor of more than 120 publications in peer-reviewed journals and conference proceedings on high-frequency circuit design and test. His main research interests are energy-efficient low-voltage circuit design, high-bandwidth and data-processing architectures and adaptive built-in-self-test systems. Dr. Margala is a member of program committees of many conferences and symposia in design and test.

October 5, 2009
Integrated Circuit Design for Neural Prostheses
Christopher Salthouse
Umass ECE Dept.
Monday, October 5, 4-5pm, ELAB 303
(CSE) and (Communications and Signal Processing)

The ability to connect a computer directly to the brain remains a dream of science fiction authors.  Cochlear Implants (CIs) are the closest thing that is commercially available.  CIs give hearing to the deaf by directly stimulating the nerves in the inner ear.  Prof.  Salthouse will explain how CIs work, how he was able to decrease the power consumption of the signal processing in CIs by more than a factor of 30, and how the building blocks of CIs are now being applied to other neural prostheses.

Prof. Christopher Salthouse joined the faculty at the University of Massachusetts on August 1st, 2009.  He is now starting the Biomedical Electronics Laboratory.  Before coming to Amherst, he was a postdoctoral fellow in the Center for Molecular Imaging Research at the Massachusetts General Hospital(MGH) for three years.  There he grew cells, imaged mice, and developed new optical sensors for biomedical imaging.  Before going to MGH, he earned a Ph.D. at the Massachusetts of Technology developing micropower mixed signal  circuits for hearing aids and cochlear implants (devices that give hearing to the deaf.)

.Back to top