Electrical and Computer Engineering
Seminars: Spring 2013
Full Duplex Wireless Systems
Speaker: Vaneet Aggarwal, AT&T Research
Time: Thursday January 31, 11:00 am
Location: LC400, Dibner Building, Five MetroTech Center, Brooklyn
Abstract
Nodes in WiFi and cellular today cannot send and receive at the same time and frequency. This is mainly because of the self-interference from the transmitter to the receiver at the same node. Recent results have shown promising designs showing that self-interference can be cancelled for use in short-range communications achieving higher throughputs. I will describe the work done by researchers at AT&T Labs in collaboration with other universities to achieve two-way communication and achieving higher throughputs. We give novel Physical and MAC layer designs which demonstrate the feasibility in WiFi type domains. Further extensions to cellular infrastructure will also be discussed.
About the Speaker
Vaneet Aggarwal received the B.Tech. degree in 2005 from the Indian Institute of Technology, Kanpur, India, and the M.A. and Ph.D. degrees in 2007 and 2010, respectively from Princeton University, Princeton, NJ, USA, all in Electrical Engineering.
He is currently Senior Member of Technical Staff-Research at AT&T Labs-Research, Florham Park, NJ. His research interests are in applications of information and coding theory to wireless systems. Dr. Aggarwal was the recipient of Princeton University's Porter Ogden Jacobus Honorific Fellowship in 2009 and AT&T Vice President Excellence Award in 2012.
Constructive and Destructive Aspets of Hardware Security for the Internet of Things
Speaker: Professor Christof Paar, University of Bochum, Germany
Time: Tuesday February 5, 11:00 am
Location: Room 9.007, Two MetroTech Center, 9th Floor, Brooklyn
Abstract
Through the steady rise of interconnected embedded systems, the vision of pervasive computing has become reality over the last few years. As part of this development, the PC-centric Internet is evolving into the Internet of Things. It turns out that securing networked embedded devices is quite a different matter from traditional Internet security. Prominent examples for embedded security include the Stuxnet virus, which has allegedly delayed the Iranian nuclear program, killer applications in the consumer area like iTunes or Amazon's Kindle, the business models of which rely heavily on IP protection, and even hacking into critical automotive functions. Perhaps not surprising, embedded security is much more closely tight to the underlying hardware than in the case of traditional network security. In this presentation I will talk about some of our research projects over the last few years in the area of embedded security.
In 1-2 generations of automobiles, car2car and car2infrastructure communication will be available for driver-assistance and comfort applications. The emerging car2x standards call for strong security features. Several 1000 incoming messages per second, the strict cost constraints, and the embedded environment makes this a challenging task. We show how an extremely high-performance digital signature engine was realized using low-cost hardware. Our signature engine is currently widely used in field trials in the USA. The next case study addresses the other end of the performance spectrum, namely lightweight cryptography. PRESENT is one of the smallest known ciphers which can be realized with as few as 1000 gates. The cipher was designed for extremely cost and power constrained applications such as RFID tags which can be used, e.g., as a tool for anti-counterfeiting of spare parts, or for other low-power applications. PRESENT has recently been adopted as ISO standard.
The "dark" side of our research deals with vulnerability analysis of embedded system. First, we show an implementation attack against a modern contactless smart card equipped with the -- cryptographically highly secure -- 3DES algorithm. The card is widely used in authentication and payment systems. The second attack breaks the bit stream encryption of current FPGAs. We were able to extract AES and 3DES keys using the power traces from a single power-up of the target device. Once the key has been recovered, an attacker can clone, reverse engineer and alter a presumingly secure hardware design.
About the Speaker
Christof Paar has the Chair for Embedded Security at the University of Bochum, Germany, and is affilated professor at the University of Massachusetts at Amherst. He co-founded, with Cetin Koc, the CHES (Cryptographic Hardware and Embedded Systems) conference. Christof’s research interests include highly efficient software and hardware realizations of cryptography, physical security, penetration of real-world systems, trusted systems and cryptanalytical hardware. He also works on real-world applications of embedded security, e.g., in cars, consumer devices, smart cards and RFID.
Christof has over 150 peer-reviewed publications and is co-author of the textbook Understanding Cryptography (Springer, 2009). He has given invited talks at MIT, Yale, Stanford University, IBM Labs, Intel and Sun Labs. He has taught cryptography extensively in industry, including courses at GTE, NASA, Motorola Research, and Philips Research. Christof is Fellow of the IEEE. He co-founded ESCRYPT Inc. - Embedded Security, a leading system provider in industrial security which was acquired by Bosch in 2012.
Integrated Wireless Senors: A Hardware Perspective
Speaker: Professor Khaled Nabil Salama, King Abdullah University of Science and Technology
Time: Tuesday February 12, 2:00 pm
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Networked systems of tiny wireless and sensing-enabled devices continue to give birth to a host of new applications that range from medical sensors for image-guided surgery, to distributed image-based surveillance of remote areas for security or environmental reasons. Such applications mandate new requirements in terms of size of the devices as well as the bandwidth required. Extreme requirements for small size packaging of the devices are obvious for many applications including biomedical ones. Fully integrated sensor modules that are capable of harvesting energy, sensing the environment and communicating with other sensors or base stations are becoming a necessity. Despite the development chips for these systems, there continues to be a need for improved implementations of micro-scale detection and processing systems for further convenience, scaling and portability. These systems would include a sensor module (mostly in mems), attached to analog front end circuitry, an analog to digital converter and a wireless communication module. We will present the research conducted at KAUST addressing many of these components. A flagship project demonstrating these concept is a single chip implantable wireless sensor system for Intraocular Pressure Monitoring (IOPM). This system-on-chip (SoC) is battery free and harvests energy from incoming RF signals, consumes 513 W of peak power and when implanted inside the eye, it can communicate over a distance of more than 15 cm.
About the Speaker
Dr. Salama received his bachelor's degree with honors from the Electronics and Communications Department at Cairo University in Egypt in 1997, and his master's and doctorate degrees from the Electrical Engineering Department at Stanford University in the United States, in 2000 and 2005 respectively. He was an assistant professor at RPI between 2005 and 2009. He joined King Abdullah University of science and technology (kaust) in January 2009 and was the electrical engineering founding program chair till August 2011. His work on CMOS sensors for molecular detection has been funded by the National Institutes of Health (NIH) and the Defense Advanced Research Projects Agency (DARPA), awarded the Stanford-Berkeley Innovators Challenge Award in biological sciences and was acquired by Lumina Inc in 2008. He is the co-author 100 papers and 10 patents on low-power mixed-signal circuits for intelligent fully integrated sensors and non linear electronics specially memristor devices. He is a senior member of IEEE.
Adaptive Multiplexing for Synchronous Multimodal Communication
Speaker: Mohamad A. Eid, University of Ottawa
Time: Thursday February 14, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Recent trends in multimedia applications strive to incorporate multi-modal media, such as audio, video, graphics, and Haptics to enhance the user’s experience. Recently, researchers have made significant progress in advanced multimedia systems by incorporating mixed reality environments and Haptics into the human computer interaction paradigm. The communication of synchronous multimodal media remains a challenge due to the different and sometimes conflicting communication requirements. In this talk, I would like to provide insights on haptic technology and synchronous multimodal interaction and propose an adaptive multiplexing framework and communication protocol for multimedia applications incorporating haptic, visual, and auditory, among other media data for non-dedicated networks. A standard description scheme for haptic-audio-visual applications, named Haptic Applications Meta Language (HAML) is described. Finally, I would like to overview my current research interests and vision for future perspectives.
About the Speaker
Mohamad Eid received the PhD in Electrical and Computer Engineering from the University of Ottawa, Canada, in 2010. He is currently an instructor and a research associate at the Division of Engineering, New York University Abu Dhabi (NYUAD), Abu Dhabi, United Arab Emirates. He was a teaching and research associate at the University of Ottawa from June 2008 until April 2012. He has won several awards for academic and research distinction including Natural Sciences and Engineering Research Council of Canada (NSERC) Award of Excellence, University of Ottawa Excellence scholarship, and Ontario Graduate Scholarship (OGS) scholarship. He is the co-author of the book: “Haptics Technologies: Bringing Touch to Multimedia”, Springers 2011, the co-chair of the 3rd International IEEE Workshop on Multimedia Services and Technologies for E-health (MUST-EH 2013), and has been a local organizing chair for Haptic-Audio-Visual Environment and Gaming (HAVE) for the years 2007, 2008, 2009, and 2010. His current research interests include Haptics and multi-modal human computer interaction, serious gaming and tangible interfaces, and biofeedback technologies.
Mixed Integer Optimization for Wireless Networks
Speaker: Lina Al-Kanj, American University of Beirut
Time: Tuesday February 19, 2:00 pm
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Applications of mixed integer optimization are in a period of rapid development due to a combination of improved modeling approaches, faster computational capabilities, and enhanced solving techniques. This enables problem modeling and solution from larger and more complex sets of alternatives which is attractive to address planning, design, and scheduling problems in the field of wireless networks. This talk discusses the utilization of mixed integer programs and solution techniques to address specific problems relevant to the design of next generation wireless networks. First, cooperative content distribution in wireless networks will be discussed for various design alternatives including optimal grouping of the mobile terminals, unicasting/multicasting communications and multihop cooperation. Second, the problem of radio network planning will be discussed including multi-technology co-siting and planning with green considerations. Optimized solutions, complexity analysis along with various polynomial time heuristic solutions will be presented and analyzed.
About the Speaker
Lina Al-Kanj received the B.E. degree in Electrical and Communications Engineering from the Lebanese University in 2005. She received her ME and PhD degrees in Electrical and Computer Engineering from the American University of Beirut in 2007 and 2012, respectively. She was a visiting PhD student at the University of Texas at Austin in 2009-2010 and a visiting Master student at Munich University of Technology in 2006. Since July 2012, she is working as a research associate at the American University of Beirut. Her research interests include network optimization, cooperative and heterogeneous wireless networks, resource allocation, and cellular radio network planning. She is the recipient of the L’Oreal-Unesco Fellowship for distinguished women in Science in 2012.
Network Optimization and Performance Evaluation in Wireless Networks, Cognitive Radio Networks, and Social Networks
Speaker: Shouling Ji, Georgia State University
Time: Thursday February 21, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
In the presentation, I will introduce my current research as well as my future research directions. First, data gathering, which includes data collection and aggregation, is one of the most fundamental operations in wireless networks. For this topic, we conducted extensive research on designing effective data gathering algorithms and analyzing the achievable network capacity under various network scenarios, e.g. randomly deployed wireless networks, probabilistic wireless networks, distributed and asynchronous wireless networks, etc. I will show our research results with emphasizing on the data collection algorithm design and capacity analysis. Second, as a promising communication paradigm, Cognitive Radio Networks (CRNs) have paved a road for secondary users to opportunistically exploit unused licensed spectrum without causing unacceptable interference to primary users. In my presentation, I will explain the characteristics and challenges in the new emerging CRN paradigm followed by introducing our works in this area. Third, social networks are important mediums for communication, information dissemination, and influence spreading. I will introduce our research on influential node set selection, network evolution analysis, and information delivery for social networks, which is motivated by applications of alleviating social problems, such as drinking, addicting to games, and information/influence spreading problems, such as promoting new products. Finally, I will also introduce my future research directions including network capacity analysis under secure communications, network optimization and performance evaluation for CRNs with considering the social behavior of primary users, and security issues in social networks.
About the Speaker
Shouling Ji is a Ph.D. Candidate in the Department of Computer Science at the Georgia State University. He received his B.S. (with Honors) and M.S. Degrees of Computer Science from Heilongjiang University, China, in 2007 and 2010, respectively. He received another M.S. Degree of Computer Science from Georgia State University in 2011. His research interests include Wireless Sensor Networks, Data Management in Wireless Networks, Cognitive Radio Networks, and Social Networks. He is now a student member of ACM, IEEE, and IEEE COMSOC. He is also the Membership Chair of the IEEE Student Branch at Georgia State.
Increasing the uplink throughput of cellular machine-to-machine communications
Speaker: Howard Huang, Bell Labs
Time: Friday February 22, 11:00 am
Location: JAB 473, Jacobs Academic Building, Six MetroTech Center, Brooklyn
Abstract
Operators are considering the use of existing cellular network infrastructure for addressing machine-type communication for sensing and monitoring applications. The communication is characterized by fixed payloads on the order of hundreds of bits and latencies on the order of hundreds of milliseconds. The conventional strategy for uplink communication (e.g., LTE) consists of two stages: a random-access stage where devices contend for service and a second stage where transmission of the actual payload occurs over assigned resources. We propose a novel single-stage strategy where users transmit their payload over the random access channel. In contrast to conventional random access transmission where users transmit over a randomly chosen subband or spreading code, users in the proposed strategy each transmit over the full bandwidth, and a receiver jointly decoders the received signal with no advance knowledge of the active users' identities. We establish the optimality of this uncoordinated strategy and show that it can double the number users transmitting 100 bits over 1MHz bandwidth with a latency of 500ms.
About the Speaker
Howard Huang received a B.S. in electrical engineering from Rice University in 1991 and a Ph.D. in electrical engineering from Princeton University in 1995. Since then, he has been a researcher at Bell Labs (Alcatel-Lucent) in Holmdel, New Jersey, currently as a Distinguished Member of Technical Staff in the Wireless Communication Theory Department. His interests are in communication theory and the system design of wireless networks.
Dr. Huang has been an active proponent of MIMO technologies in 3GPP standards, representing Bell Labs when MIMO was first proposed in 2000 for UMTS. He worked as the rapporteur for the 3GPP MIMO work item and later made contributions on MIMO for the LTE and LTE-Advanced standards. He served as a guest editor of two issues of the IEEE Journal of Selected Areas in Communications-- one on next-generation MIMO wireless networks and another on cooperative communications in MIMO cellular networks. Dr. Huang has taught as an adjunct professor at Columbia University and is a co-author of the book MIMO Communication for Cellular Networks (published by Springer 2011). He holds over a dozen patents related to wireless communications and is a Senior Member of the IEEE.
Understanding Photos of People Using Social Context
Speaker: Professor Tsuhan Chen, Cornell University
Time: Friday March 1, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Our eyes and brains are fine-tuned to view and analyze images of people. When we view an image of people, we can make judgment very effectively. For example, we can easily come up with demographic descriptions of people in the image. We can also answer other questions related to the activities of, and relationships between, people in the image. All that reasoning happens not only because of what our eyes see, but also how our brain draws "prior," or context, from experiences. In this talk, we will present some recent discovery in how computer algorithms can be developed to do the same as our brain, that is, to use social context to understand photos of people. This approach has a lot of potential, as the number of photos shared by users of social networks increases exponentially.
About the Speaker
Tsuhan Chen has been with Cornell University, Ithaca, New York, since January 2009, where he is the David E. Burr Professor of Engineering and Director of the School of Electrical and Computer Engineering. From October 1997 to December 2008, he was with the Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania, as Professor, and as Associate Department Head in 2007-2008. From August 1993 to October 1997, he worked at AT&T Bell Laboratories, Holmdel, New Jersey. He received the M.S. and Ph.D. degrees in electrical engineering from the California Institute of Technology, Pasadena, California, in 1990 and 1993. He received the B.S. degree in electrical engineering from the National Taiwan University in 1987. He received the Benjamin Richard Teare Teaching Award in 2006, and the Eta Kappa Nu Award for Outstanding Faculty Teaching in 2007. He was elected to the Board of Governors of IEEE Signal Processing Society, 2007-2009, and a Distinguished Lecturer of IEEE Signal Processing Society, 2007-2008. In 2012, he was elected as the Vice President of ECE Department Head Association, and will serve as the President in 2013. He is a Fellow of IEEE.
Acoustic Resonators on CMOS as Label-free Chemical Sensors
Speaker: Matthew L. Johnston, Columbia University
Time: Monday March 4, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Merging chemical and biological sensors with silicon integrated circuits has the potential to push complex electronics into low-cost, portable detection platforms. This greatly simplifies system-level instrumentation and extends the reach and functionality of pointof-care technologies. One such class of sensor, the piezoelectric thin-film bulk acoustic resonator (FBAR), has a micron-scale size and low gigahertz frequency range that is ideally matched with modern complementary metal-oxide-semiconductor (CMOS) technologies. This device is analogous to a quartz crystal microbalance, where selective analyte attachment causes a proportional shift in resonant frequency. An FBAR sensor enables label-free detection of analytes in real time, and CMOS integration overcomes the measurement complexity and equipment cost normally required for quantitative detection with acoustic resonators.
In this talk, I will present a piezoelectric FBAR array on CMOS for physical sensing applications. We have developed a monolithic fabrication method that enables the construction of unreleased, high-Q resonators directly on the top surface of a standard CMOS substrate. As a demonstrated application, I will present results of quantitative volatile organic compound (VOC) sensing using polymer-functionalized FBAR-CMOS sensors. Finally, I will discuss ongoing efforts to extend this platform to a clinical application in radiation biodosimetry, as well as outline potential applications in environmental monitoring, clinical diagnostics, and RF microelectronics.
About the Speaker
Matt Johnston is a postdoctoral research associate in the Bioelectronic Systems Lab at Columbia University, where his research focuses on IC-enabled sensor platforms for chemical and biological applications. Matt received the B.S. degree in electrical engineering from the California Institute of Technology (Caltech) in 2005, and the M.S. and Ph.D. degrees in electrical engineering from Columbia University in 2006 and 2012, respectively. Matt was previously co-founder and Head of Research at Helixis, a Caltech based startup company developing low-cost, real-time PCR systems that was acquired by Illumina in 2010. Matt has also worked with Cavium Networks and The Aerospace Corporation. His current research interests include biosensors and bioelectronics, microfluidics, and point-of-care technologies for medical monitoring and clinical diagnostics.
Sensing Driver Phone Use to Reduce Driver Distraction
Speaker: Professor Yingying Chen, Stevens Institute of Technology
Time: Thursday March 7, 11:00 am
Location: LC400, Dibner Building, Five MetroTech Center, Brooklyn
Abstract
Distinguishing driver and passenger phone use is a building block for a variety of mobile applications. And it’s greatest promise lies in helping reduce driver distraction. Cell phone distractions have been a factor in high-profile accidents and are associated with a large number of automobile accidents. In this talk, I will describe a system that lets mobile phones detect whether they are used by a driver. Our detection system addresses the fundamental problem of distinguishing between a driver and passenger phone use. It leverages the existing car stereo infrastructure, in particular, the speakers and Bluetooth network. We utilizes an acoustic ranging based approach to estimate the phone’s distance from the car’s center, from which a passenger or driver classification can be made. We have prototyped and demonstrated our system on Android phones. Through extensive experiments with various types of phones in two different cars, we find that Our customized beeps for acoustic ranging are robust to background sounds such as music and wind, and the signal processing did not require excessive computational resources on smartphones. In spite of the cars’ heavy multi-path environment, our approach can achieve a classification accuracy of about 95%. This work has received the Best Paper Award from ACM International Conference on Mobile Computing and Networking (MobiCom) 2011. I will also sketch our recent study towards this direction by only using sensors on smartphones.
About the Speaker
Yingying (Jennifer) Chen is an associate professor in the Department of Electrical and Computer Engineering at Stevens Institute of Technology. Her research interests include cyber security and privacy, wireless networks, mobile social networks and pervasive computing. She received her Ph.D. degree in Computer Science from Rutgers University. She has coauthored the book Securing Emerging Wireless Systems and published extensively in journal and conference papers. Prior to joining Stevens Institute of Technology, she was with Alcatel-Lucent. She received the IEEE Outstanding Contribution Award from IEEE New Jersey Coast Section each year 2005-2009. She is the recipient of the NSF CAREER award on wireless security and Google Research Award. She received Stevens Board of Trustees Award for Scholarly Excellence. She received New Jersey Inventors Hall of Fame Innovators Award 2012. She is also the recipient of the Best Paper Award from ACM International Conference on Mobile Computing and Networking (MobiCom) 2011 and International Conference on Wireless On-demand Network Systems and Services (WONS) 2009, as well as the Best Technological Innovation Award from the International TinyOS Technology Exchange 2006. Her research has been reported by numerous media outlets including the Wall Street Journal, MIT Technology Review, Inside Science, Tonight Show with Jay Leno, NPR, and CNET.
Heat-Diffusion Approaches for 3D Computer Vision
Speaker: Yi Fang, Purdue University
Time: Thursday March 7, 2:00 pm
Location: LC400, Dibner Building, Five MetroTech Center, Brooklyn
Abstract
Recent developments in data acquisition techniques have resulted in a rapid growth in the number of available three dimensional (3D) models across areas as diverse as engineering, medicine and biology. Researchers are regularly interested in interpreting the 3D shape of such models according to their intrinsic geometric attributes. The effective and efficient interpretation of 3D models is often challenged with the prevalence of non-rigidity within the shapes, the corruption of the shapes due to the presence of geometric noise, and the availability of a large volume of 3D models in innumerable databases. My work is concentrated on the development of a novel framework for 3D shape analysis, such as shape matching, segmentation, and retrieval, based on the effective utilization of the heat diffusion concept. The novelty of this framework is derived from an analogy between the process of 3D shape interpretation and that of heat transfer. The approaches exploit the intelligence of heat as a global structure-aware message that traverses across a meshed surface and is capable of exploring the intrinsic geometric features of the shape. I have demonstrated the performance of several heat-driven approaches for efficient non-rigid 3D shape registration, robust segmentation of 3D models, and efficient retrieval of 3D models with applications in engineering, medicine and biology. The experimental results indicate that heat-driven approaches are able to reveal the interpretations of 3D shape in a highly robust fashion, independent of any reference to prior knowledge, and in a manner consistent to human perception. In addition, the heat-driven approaches are very general and have great potential for applications to a broad range of research fields, for example, medical image processing, biological networks, social networks and semantic analysis of documents.
About the Speaker
Yi Fang received his Bachelor's and Master's degree in Engineering from Xi’an Jiao tong University, China, in 2003 and 2006, respectively. He received his PhD degree in Engineering from Purdue University, West Lafayette, USA on December, 2011. He worked as research intern in Siemens Corporate Research on 3D medical image processing. He then joined Riverain Technologies, a leader and technology innovator in the health care industry and beyond, as a Senior Research Scientist. He is now a Senior Staff Scientist in the Department of Electrical Engineering and Computer Science at Vanderbilt University. His current research interests are in computer graphics, computer vision, image processing, machine learning and their applications to multiple disciplines as diverse as engineering, medicine, biology and social science. He has co-authored 16 refereed papers in journals and top tier conferences, of which he is the first author in 9. Some of his works have been widely reported by both national and international media, such as Purdue Newsroom, Sciencedaily and Yahoo!.
Designs of Flash and Associative Memories
Speaker: Eitan Yaakobi, California Institute of Technology
Time: Monday March 11, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Flash is the leading technology for non-volatile memory today. However, flash suffers from an asymmetry between cell programming and cell erasing; while it is easy to increase a cell charge, reducing its charge requires erasing a large block of cells. Block erasures are not only time-consuming, but also degrade the lifetime of the memory. Rewriting algorithms are one of the efficient approaches to mitigate the lifetime constraint and improve endurance. I will present recent advances in rewriting algorithms as well as generalizations of the work of Rivest and Shamir on Write Once Memories.
The second part of this talk considers associative memories, where the focus is on efficient management and retrieval of information. Our approach is inspired by our understanding of the information processing in the brain. The human brain stores information by associations and this organization leads to efficient data retrieval. I will present results related to the concept of uncertainty in associative memories and show its connection to Levenshtein’s sequences reconstruction problem.
About the Speaker
Eitan Yaakobi is a postdoctoral researcher in the department of Electrical Engineering at the California Institute of Technology, where he works with Prof. Shuki Bruck. He is also affiliated with the Center for Magnetic Recording Research at the University of California, San Diego. He received a PhD. degree in Electrical and Computer Engineering at the University of California, San Diego, under the supervision of Prof. Paul Siegel, Prof. Alexander Vardy, and Prof. Jack Wolf. He received the Marconi society young scholar award in 2009 and was a recipient of the Intel Ph.D. fellowship in 2010-2011. His research interests include information and coding theory with applications to non-volatile memories, associative memories, data storage and retrieval, and voting theory.
A Game-Theoretic Approach for Resilient, Robust and Secure Control of Cyber-Physical Systems
Speaker: Quanyan Zhu, University of Illinois at Urbana-Champaign
Time: Monday March 11, 2:00 pm
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
With its rich set of conceptual, analytical and algorithmic tools, game theory has emerged as providing a versatile and effective framework for addressing issues of robustness, resilience, and security (RRS) in modern critical infrastructures. Such systems are composed of many interacting human, cyber and physical components at multiple layers. Addressing issues of RRS will require a divide-and-conquer approach, and at the same time a holistic system viewpoint. The talk will first introduce game- and control-theoretic approaches for modeling multi-layer and multi-agent interactions in cyber-physical systems. It will then present the recently developed resilient control theory for efficient cyber and physical system integrations for achieving cyber security against attackers and robustness of physical system against noise and disturbances. The talk will also discuss the games-in-games principle and multi-resolution game theory to address strategic decision-making residing at multiple layers of the cyber-physical system. Specific examples will be drawn from communication networks and smart energy systems for illustration of these concepts.
About the Speaker
Quanyan Zhu is a PhD candidate at the Department of Electrical and Computer Engineering and the Coordinated Science Laboratory (CSL) at University of Illinois at Urbana-Champaign (UIUC). He has received his Master of Applied Science in Electrical Engineering from University of Toronto, and Bachelor of Engineering in Honors Electrical Engineering from McGill University. He has been a visiting researcher at University of Waterloo, University of Avignon, University of Houston, INRIA-Sophia Antipolis, Idaho National Laboratory, SUPELEC, University of Washington and Chinese Academy of Mathematics and System Science. He is a recipient of NSERC Canada Graduate Scholarship, University of Toronto Fellowship, Ernest A. Reid Fellowship and Mavis Future Faculty Fellowships. He is a recipient of the best track paper award at the 4th international symposium on resilient control systems (ISRCS). He is the organizer of the resilient control system tutorial at CPSWEEK 2012, the TPC Chair of the 1st and 2nd INFOCOM workshop on communications and control on smart energy systems (CCSES), and the organizer of the 1st and 2nd Midwest Workshop on Control and Game Theory (WCGT).
Healthcare Promotion: End-to-End Research in Wireless Health
Speaker: Professor Wenyao Xu, University of California, Los Angeles
Time: Tuesday March 12, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Solving healthcare related problems is one of the grand challenges in the 21st century. In many nations, the need to improve existing medical and healthcare services is becoming increasingly important due primarily to the growing population and ageing society. The rapid advancement of sensors, electronics and wireless communication technology has brought rise to a new research field called Wireless Health. This field aims to deliver solutions in healthcare that can not only be resolved by pieces of engineering work, but require the efforts of many people in Engineering, Computer Science and Medicine, resulting in 'End-to-End Research in Wireless Health'.
In this talk, I will introduce my research experiences in Wireless Health. I highlight how to team up with medical professionals to solve real healthcare problems. Through a concrete example, Smart Bedsheet, I elaborate on our end-to-end work of developing healthcare solutions to enable pressure ulcer (bedsore) reduction, including identifying problems, proposing solutions, developing systems and clinical verification. In addition, I also introduce other projects that address important health challenges, such as fall prevention, obesity control and rehabilitation.
About the Speaker
Wenyao Xu is a Ph.D candidate at the University of California, Los Angeles, supervised by Majid Sarrafzadeh in the Wireless Health Institute. He obtained his B.S. and M.S. from Zhejiang University in 2006 and 2008, respectively. His research interests are in the area of embedded systems, computational modeling and algorithm design. During his Ph.D study, he focused on new sensing and computing technologies for healthcare applications. Wenyao received best demo paper awards in ACM Wireless Health Conference 2011 and Body Sensor Network Conference 2012. He holds five U.S. patents, which are licensed to nationally renowned bio-medical device companies. He is a co-founder of Medisens Wireless Inc., a start-up in San Jose, focused on commercializing several of his research products.
Modeling and Optimizing Heterogeneous Cellular Network Capacity
Speaker: Professor Jeffrey Andrews, University of Texas at Austin
Time: Tuesday March 12, 2:00 pm
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Given the ever increasing heterogeneity, density, and irregularity of cellular networks, new models are needed for understanding the statistics of SINR, and rate. Unlike in conventional cellular networks, SINR does not map directly to rate, because the rate is largely a function of the congestion on the base station. We begin by reviewing our proposed spatial model and SINR results, before moving onto attempting to understand the optimal rate distribution in HetNets. To do so, the rules used for associating mobile devices with base stations need to be revisited from first principles. We explore this challenging optimization problem from two angles; the first being a fairly traditional optimization approach with several relaxations. Interestingly, we show numerically that the simple "cell range expansion" approach advocated by Qualcomm and adopted by 3GPP is near-optimal despite its simplicity. The second approach is an approximate statistical approach based on stochastic geometry. In both cases, we observe interesting and largely compatible trends, indicating the great benefits available from intelligent association.
About the Speaker
Jeffrey Andrews is a Professor in the Department of Electrical and Computer Engineering at the University of Texas at Austin. His expertise is on both theoretical and practical aspects of wireless systems and networks, recently focused on topics in heterogeneous cellular networks such as femtocells. He has held industry positions at Qualcomm (1995-97) and Intel (1994), as well as consulting for Verizon, the WiMAX Forum, Apple, Intel, Microsoft, Clearwire, Sprint, and NASA.
Dr. Andrews is the co-author of two books, the popular Fundamentals of WiMAX (Prentice-Hall, 2007) and Fundamentals of LTE (Prentice-Hall, 2010), as well as over 250 publications, many highly cited, and several patents. He is an IEEE Fellow, has received five IEEE Best Paper awards, the National Science Foundation CAREER award, and holds a Ph.D. in Electrical Engineering from Stanford University.
Compression and Modern Data Processing
Speaker: Thomas Courtade, Stanford University
Time: Wednesday March 13, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
At first glance, modern applications of data processing -- such as clustering, querying, and search -- bear little resemblance to the classical Shannon-theoretic problem of lossy compression. However, the ultimate goal is the same for modern and classical settings; both demand algorithms which strike a balance between the complexity of the algorithm output and the utility that it provides. Thus, when we attempt to establish fundamental performance limits for these "modern" data processing problems, elements of classical rate distortion theory naturally emerge.
Inspired by the challenges associated with extracting useful information from large datasets, I will discuss compression under logarithmic loss. Logarithmic loss is a penalty function which measures the quality of beliefs a user can generate about the original data upon observing the compressor's output. In this context, we characterize the tradeoff between the degree to which data can be compressed and the quality of beliefs an end user can produce. Notably, our results for compression under logarithmic loss extend to distributed systems and yield solutions to two canonical problems in multiterminal source coding.
I will also briefly discuss recent work on compression for identification, where we seek to compress data in a manner that preserves the ability to reliably answer queries of a certain form. This setting stands in stark contrast to the traditional compression paradigm, where the goal is to reproduce the original data (either exactly or approximately) from its compressed form. Under certain assumptions on the data sources, we characterize the tradeoff between compression rate and the reliability at which queries can be answered.
About the Speaker
Thomas Courtade received the B.S. degree in Electrical Engineering from Michigan Technological University in 2007, and the M.S. and Ph.D. degrees in Electrical Engineering from UCLA in 2008 and 2012, respectively. In 2012, he was awarded the Inaugural Postdoctoral Research Fellowship through the Center for Science of Information. He currently holds this position, and resides at Stanford University. His recent honors include a Distinguished Ph.D. Dissertation award and an Excellence in Teaching award from the UCLA Department of Electrical Engineering and a Best Student Paper Award at the 2012 International Symposium on Information Theory.
Adaptive Load Management: Scheduling And Coordination Of Demand Resources In Power Systems
Speaker: Jhi-Young Joo, Carnegie Mellon University
Time: Wednesday March 13, 2:00 pm
Location: RH505, Rogers Hall, 6 MetroTech Center, Brooklyn
Abstract
Demand response refers to techniques that manage end-users' electricity consumption in order to help the power system operate in a more cost-efficient and reliable way. It is becoming more important with increasing renewable and distributed energy resources incorporated into the system. Our proposed demand response framework, namely Adaptive Load Management, provides a comprehensive structure for demand response by formulating the complex power system objective as many sub-problems of diverse supply and demand entities in the system over multiple time horizons. In this talk, I will focus on the short-term scheduling of supply and demand resources, with an emphasis on managing flexible demand resources of small end-users. The difficulty of this problem comes from the uncertainty of loads and supply, limitations on communication and information exchange among a large number of supply and demand entities, and modeling different values of electric energy seen by diverse end-users/loads. I will address how we tackle these issues, and present simulation results that demonstrate how our methodology works.
About the Speaker
Jhi-Young Joo is a Ph.D. candidate in the Department of Electrical and Computer Engineering at Carnegie Mellon University. She received her B. Eng. and M. Eng. Degrees from the School of Electrical and Computer Engineering at Seoul National University, Korea in 2005 and 2007, respectively. She co-founded Carnegie Mellon Electric Energy Club in 2009 and served as the treasurer and the vice president in the following two years. Her research interests include modeling and optimization of power systems and energy markets.
Computable Performance Analysis of Sparse Recovery with Applications
Speaker: Professor Arye Nehorai, Washington University in St. Louis
Time: Thursday March 14, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
The last decade has witnessed burgeoning developments in the reconstruction of signals based on exploiting their low-dimensional structures, particularly their sparsity, block-sparsity, and low-rankness. The reconstruction performance of these signals is heavily dependent on the structure of the operating matrix used in sensing. The quality of these matrices in the context of signal recovery is usually quantified by the restricted isometry constant and its variants. However, the restricted isometry constant and its variants are extremely difficult to compute.
We present a framework for analytically computing the performance of the recovery of signals with sparsity structures. We define a family of incoherence measures to quantify the goodness of arbitrary sensing matrices. Our primary contribution is the design of efficient algorithms, based on linear programming and second order cone programming, to compute these incoherence measures. As a by-product, we implement efficient algorithms to verify sufficient conditions for exact signal recovery in the noise-free case. The utility of the proposed incoherence measures lies in their relationship to the performance of reconstruction methods. We derive closed-form expressions of bounds on the recovery errors of convex relaxation algorithms in terms of these measures. The second part of the talk applies the developed theory and algorithms to the optimal design of an OFDM radar system with multi-path components.
About the Speaker
Arye Nehorai is the Eugene and Martha Lohman Professor and Chair of the Preston M. Green Department of Electrical and Systems Engineering at Washington University in St. Louis (WUSTL). Under his leadership as department chair, the undergraduate enrollment has tripled in the last four years.Earlier he was a faculty member at Yale University and the University of Illinois at Chicago. In 2001 he was named University Scholar of the University of Illinois.He received the B.Sc. and M.Sc. degrees from the Technion, Israel, and the Ph.D. from Stanford University, California.Dr. Nehorai has served as Editor-in-Chief of the IEEE Transactions on Signal Processing during the years 2000 to 2002. From 2003 to 2005 he was Vice President (Publications) of the IEEE Signal Processing Society (SPS), Chair of the Publications Board, and member of the Executive Committee of this Society. He was the Founding Editor of the special columns on Leadership Reflections in the IEEE Signal Processing Magazine from 2003 to 2006.Dr. Nehorai received the 2006 IEEE SPS Technical Achievement Award and the 2009 IEEE SPS Meritorious Service Award. He was elected Distinguished Lecturer of the IEEE SPS for the term 2004 to 2005. He co-authored three journal papers that received best paper awards from the IEEE Signal Processing Society. He was also co-author of five best paper awards in student competitions at international conferences.He is a Fellow of the Institute of Electrical and Electronics Engineers (IEEE), the Royal Statistical Society, and the American Association for the Advancement of Science (AAAS).
Architectural Support for Securing Cloud Servers
Speaker: Jakub Szefer, Princeton University
Time: Thursday March 15, 11:00 am
Location: JAB473, Jacobs Academic Building, Six MetroTech Center, Brooklyn
Abstract
Cloud computing is becoming a dominant computing paradigm. However, most cloud computing services are built using commodity systems not designed to handle the variety of threats present in this utility-like computing model. Users' concerns and surveys of hypervisor vulnerabilities have motivated our research on securing virtual machines, in particular we focus on protections from a malicious or compromised hypervisor. We have defined hypervisor-free virtualization, realized in the NoHype architecture, which aims to eliminate the need for active hypervisor when the virtual machines run. Our key insight is to use hardware virtualization features, originally deigned for performance reasons, to remove the hypervisor attack surface and securely isolate the virtual machines. We also defined hypervisor-secure virtualization, realized in the HyperWall architecture, which further improves virtual machine security while providing more functionality over NoHype. The HyperWall architecture allows an untrusted commodity hypervisor to manage the system while the virtual machines are protected from it. Our key contribution is a special new feature we introduced: the hardware-only accessible DRAM for storing the protections. To improve confidence in the security of the design, we recently proposed a novel security verification methodology, and applied it to component interactions and protocols of HyperWall. By designing and verifying such architectures for secure cloud computing, we can enable more users to enjoy the benefits of cloud computing and be able to securely process sensitive code and data in virtual machines running on cloud servers – even if attackers can gain hypervisor-level privileges.
About the Speaker
Jakub Szefer’s research interests are at the intersection of computer architecture and computer security. His recent work focuses on securing cloud computing, even if the hypervisor running on the cloud servers is compromised. He received B.S. degree with highest honors in Electrical and Computer Engineering from University of Illinois at Urbana-Champaign in 2006, a M.A. in Electrical Engineering rom Princeton University in 2009, and expects his Ph.D. also in Electrical Engineering from Princeton University in early 2013. He is part of the Princeton Architectural Lab for Multimedia and Security (PALMS) led by Prof. Ruby B. Lee. In addition to research, he enjoys teaching and has won two outstanding TA awards and the Wu Prize for Excellence.
Data Analytics for Smart Grid: Spatio-temporal Wind Power Analysis and Synchrophasor Data Mining
Speaker: Miao He, Arizona State University
Time: Wednesday March 27, 11:00 am
Location: RH505, Rogers Hall, 6 MetroTech Center, Brooklyn
Abstract
Future smart grids can potentially generate massive amounts of new detailed data from widely deployed measurement devices at all domains (bulk generation, transmission, distribution, end-user, etc). One central issue in managing the complex, diverse and distributed data under increasingly dynamic and uncertain conditions of smart grid is the effective extraction of relevant information that can enhance situational awareness and decision making. In this presentation, two specific topics will be discussed, i.e., spatio-temporal analysis for wind farm generation forecast and synchrophasor data mining for online power system dynamic security assessment (DSA). Specifically, the first part will begin with a brief introduction to the spatial and temporal dynamics of wind farm generation observed from extensive measurement data. Then, a general distributional forecast model that can be used in stochastic unit commitment and dispatch problems will be presented. In the second part of the presentation, a data mining-based DSA approach, which is robust to the uncertainty and dynamics of both the cyber and the physical systems of smart grid, will be presented. Results from several case studies using realistic power system models and wind farm generation data, together with the insight in applying both data-driven and model-based tools for data analytics of smart grid, will be discussed.
About the Speaker
Miao He received his B.E. degree from Nanjing University of Posts and Telecommunications, China, in 2005 and his M.E. degree from Tsinghua University, China, in 2008, both in Electrical Engineering. Currently, he is a Ph.D. candidate in the School of Electrical, Computer and Energy Engineering at Arizona State University. His research interests include data analytics of smart grid, renewable generation forecast and integration, wide-area monitoring and protection of power systems. He is a student member of IEEE, IEEE Power and Energy Society (PES) and IEEE Communications Society (ComSoc).
Nonlinear Control of Delay and PDE Systems: Methods and Applications
Speaker: Nikolaos Bekiaris-Liberis, University of California at San Diego
Time: Thursday March 28, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Delay and PDE systems are ubiquitous in applications. The distributed nature of such processes creates challenges in control. I will present my recent results on the nonlinear control of delay systems with nonconstant delays and on the nonlinear control of coupled PDE systems, using the concepts of nonlinear predictor feedback and infinite-dimensional backstepping transformation. I will introduce a methodology for the control of gas emissions in IC engines using a PDE-based approach. I will explain how the control and analysis tools, that I have developed, can be applied to four major technological problems--control over networks, 3-D printing, oil drilling, and cooling systems. I will conclude by addressing some additional potential future research areas.
About the Speaker
Nikolaos Bekiaris-Liberis is a PhD candidate at UC, San Diego, graduating in Spring 2013. He has coauthored 12 journal papers and the upcoming SIAM book "Nonlinear Control under Nonconstant Delays." Bekiaris-Liberis taught tutorials on control of nonlinear delay systems at the 2012 IEEE Conference on Decision and Control and within the European Embedded Control Institute in 2013. He was guest editor for a special issue in the International Journal of Adaptive Control and Signal Processing. Bekiaris-Liberis was finalist for the student best paper award at the 2010 ASME Dynamic Systems and Control Conference. His interests are in distributed parameter systems, nonlinear control, and applications in automotive engines and catalysts, additive manufacturing and manufacturing processes, oil drilling and production, networks, traffic systems, robotics, and energy systems.
Emerging Sensor Systems
Speaker: Ross Walker, Stanford University
Time: Friday March 29, 11:00 am
Location: RH304, Rogers Hall, 6 MetroTech Center, Brooklyn
Abstract
Sensor systems are a ubiquitous part of modern life and make huge impacts on how we deal with disease and injury, how we communicate and travel, and how we understand our environment and ourselves. Miniaturized sensors fueled by the MEMS revolution and progress in integrated circuit technology have paved the way for small, cheap, and highly effective sensor systems. Quantum physics and materials science are bringing in the next wave of sensors that can be made orders of magnitude smaller and are capable of new modes of information transduction. Two emerging sensor systems outlined below will be discussed, and future research directions will be presented.
Progress in head-mounted neural recording platforms at Stanford University, reported in the literature as the ‘Hermes’ project series, has motivated the research and design of a new front-end ASIC solution for broadband acquisition of 96 channels of neural data from chronically implanted “Utah” intracortical microelectrode arrays (UEA). The proof-of-concept chip consumes 6.4mW from 1.2V while occupying 5mm × 5mm in 0.13µm CMOS, and enables basic neuroscience as well as neural prosthetics research by providing high fidelity simultaneous recordings from all available UEA channels with low power consumption and a compact form factor. The front-end IC and functional HermesE system prototype will be discussed.
A new label-free electronic biosensing technique is explored, based on quantum information transduction of a solvated analyte’s chemical composition. The technique combines the advantages of conventional label-free and mass spectrometry technologies by leveraging the physics of non-adiabatic, quantum, charge-transfer-related transitions at nanofabricated electrochemical interfaces. The quantum transition-based current is a rich new source of information about chemical composition, upon which we are creating a new biomolecular sensing platform for detection of biothreat agents such as botulinum toxin.
About the Speaker
Ross Walker is a graduate student at Stanford University. He received the B.S. degree in electrical engineering and the B.S. degree in computer science from the University of Arizona, Tucson, in 2005. In 2007, he received the M.S. degree in electrical engineering from Stanford University, Stanford, CA. Since 2007 he has been working toward the Ph.D. degree at Stanford University. He recently defended his thesis research and is set to graduate in June 2013. From 2003-2004 he held internships at IBM and National Semiconductor, both in Tucson, AZ. In 2006 he held an internship at Linear Technology, Milpitas, CA. His research interests include mixed signal integrated circuit design with emphasis on sensor interfacing, signal processing, and biomedical applications. Ross has worked on biomedical optical imaging systems, direct neural interface systems, quantum biomolecular transducers, and other sensor related projects.
Combinatorial Regression Techniques for Sparse Processing
Speaker: Amin Khajehnejad, California Institute of Technology
Time: Monday April 1, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
I will talk about the adoption of some combinatorial techniques in the reconstruction of high dimensional sparse data sets and low rank matrices from lower dimensional linear projections. The proposed techniques facilitate the decoding of (analog) sparse signals with extremely large dimensions in times substantially less time than the traditional convex optimization and greedy based approaches. I will mention specific statistical inference applications where the proposed methods can be used for dimensionality reduction and learning. Examples of such application are in neighbour discovery in wireless ad-hoc networks, market basket analysis, prediction of financial market events, political ranking and digital health.
About the Speaker
Amin Khajehnejad finished his PhD in electrical engineering from California Institute of Technology in June 2012. His areas of expertise are statistical signal processing, optimization, machine learning, dimensionality reduction techniques, information theory and coding. His thesis was titled "combinatorial regression and improved basis pursuit for sparse recovery", for which he received the Wiltz prize for outstanding research in electrical engineering. He obtained MS and BS degrees in electrical engineering from Caltech and University of Tehran in 2009 and 2007, respectively. In addition, he has had consulting appointments with various firms and research labs such as Lyric Semiconductor Inc., NEC Laboratories America, Proteus Digital Health, D.E Shaw & Co. and some other financial firms.
High-Density Conformal Neural Interface
Speaker: Liang Guo, Massachusetts Institute of Technology
Time: Tuesday April 2, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Numerous applications in neuroscience research and neural prosthetics, such as retinal prostheses, spinal-cord surface stimulation for prosthetics and electrocorticogram (ECoG) recording for epilepsy detection, involve electrical interaction with soft excitable tissues using a surface stimulation and/or recording approach. These applications require an interface that is able to set up electrical communications with a high throughput between electronics and the excitable tissue and that can dynamically conform to the shape of the soft tissue. Being a compliant and biocompatible material with mechanical impedance close to those of soft tissues, polydimethylsiloxane (PDMS) offers excellent potential as the substrate material for such neural interfaces. However, fabrication of electrical functionalities on PDMS has long been very challenging.
My talk will focus on the development of PDMS-based high-density conformal neural interfaces and their application as an epimysial (i.e., on the surface of muscle) interface to neural prosthetics and theragnosis. Specifically, challenges associated with the microfabrication of PDMS-based stretchable electronics, including high-density interconnect patterning, multilayer implementation and integrated packaging, will be addressed; and efforts on example medical applications, including a prosthesis for unilateral vocal cord paralysis (UVCP) and a theragnostic system for promoting peripheral nerve repair, will be described. I will conclude on perspectives on future high-throughput neural interfacing technology.
About the Speaker
Liang Guo received his B.E. degree in Biomedical Engineering from Tsinghua University, Beijing in 2004 and his Ph.D. degree in Bioengineering from the Georgia Institute of Technology, Atlanta, GA in 2011. His Ph.D. dissertation is on "High-density stretchable microelectrode arrays: an integrated technology platform for neural and muscular surface interfacing". He is presently a Postdoctoral Associate in the Laboratory of Professor Robert S. Langer at the Massachusetts Institute of Technology, Cambridge, MA. His primary research interests are in neural interfacing technology and biological circuits engineering as applied to neural prosthetics.
Fascinating Applications of Metamaterials and Plasmonics: Cloaking, Sensing, Energy Harvesting and Wireless Communication
Speaker: Pai-Yen Chen, University of Texas at Austin
Time: Wednesday April 3, 11:00 am
Location: LC400, Dibner Building, Five MetroTech Center, Brooklyn
Abstract
Metamaterials are artificial composite materials engineered to have novel electromagnetic properties, unavailable in nature and radically different from their constituent components. Plasmonics has further enriched and enhanced the field of metamaterials, opening new possibilities to manipulate and confine light at nanoscale dimensions, unthinkable only a few years ago. In my talk, I will describe my recent research efforts on plasmonic materials and metamaterials and their practical uses in energy harvesting, sensing, communication and wireless health systems based on novel electromagnetic phenomena and electronic physics in the spectral range from radio frequencies (RF) and microwaves, terahertz (THz) to visible light. I will discuss theory and practice of how the plasmonic materials and metamaterials can ultimately manipulate the relevant electromagnetic constitutive parameters, including permittivity, permeability, nonlinear susceptibility and conductivity, to offer new promises in nanoscale nonlinear optics and information processing, highly-efficient solar and thermal energy harvesting and conversion systems, and metamaterial-based/-inspired cloaks and electrically-small antennas used for enhancing the sensitivity and signal-to-noise ratio in future RF wireless communication and sensor networks. As an extreme case of signal manipulation at the “atomic” scale, I will discuss how the gate-tunable surface plasmon polaritons in graphene nanodevices may enable the THz frequency-configurable antennas and beam-steerable phased arrays. I will conclude my talk discussing the integration of graphene-based THz frequency synthesizers, antennas and circuit components to realize “all-graphene” THz transceivers and sensors of great interest in data transformation, sensing, actuation and communications of nanosystem.
About the Speaker
Pai-Yen Chen is currently a PhD Candidate under the supervision of Prof. Andrea Alù in the Department of Electrical and Computer Engineering at The University of Texas at Austin. His scientific research is in multidisciplinary areas of physical and wave electronics in the spectral range from microwaves, terahertz (THz) to visible light. His doctoral work mainly focuses on metamaterials, nanomaterials and plasmonics, as well as their applications in wireless communications, radar and sensors, electronic warfare, thermal and solar energy harvesting and conversion, biological and medicine detection. Prior to joining UT Austin (2006-2009), he studied extensively vacuum nanoelectronics and semiconductor device modeling, parameter extraction, characterization and fabrication techniques at National Nano Device Laboratories (NDL), Taiwan. Mr. Chen has published approximately 35 papers in peer-reviewed journals (3 journal coverages), 28 conference proceedings, and 2 book chapters. He is a reviewer of over 10 scientific journals.
Pai-Yen Chen received his M.S. and B.S. degrees in Electro-optical Engineering and Mechanical Engineering from National Chiao Tung University, Taiwan, in 2000 and 2004, respectively. His honors and awards include the 2005 United Microelectronics Corp. (UMC) Scholarship, Chinese Phi Tau Phi Honorable Member (officially nominated in 2006), 2009 Taiwanese Ministry of Education Study Abroad Award, 3rd prize Student Contest Award in Metamaterials'2011, 3rd prize Student Contest Award in 2012 USNC-URSI National Radio Science Meeting, Finalist and Honorable Mention Student Contest Award in 2010 and 2011 IEEE Antennas and Propagation Symposium. In 2012, he received the Donald D. Harrington Dissertation Fellowship, which is the most prestigious fellowship to graduate students bestowed by the Harrington Society at University of Texas at Austin.
Model-Based Reconstruction for Accelerated Magnetic Resonance Imaging
Speaker: Daniel S. Weller, University of Michigan
Time: Thursday April 4, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Compared to other medical imaging methods, magnetic resonance imaging (MRI) is differentiated by excellent soft tissue contrast and use of non-ionizing electromagnetic radiation. By accelerating MR image acquisition, the throughput, and hence, the affordability and comfort to patients can be increased, and the temporal resolution of dynamic and functional MRI can be improved. In this talk, I describe how my research into model-based reconstruction facilitates accelerating MRI.
To begin, I explore using sparsity to improve the GRAPPA method for accelerated MRI reconstruction with parallel receiver coils. I describe a post-processing method that effectively denoises the reconstruction, while preserving the acquired data. Another approach regularizes calibration of the GRAPPA method to promote sparsity in the output, improving reconstruction quality when insufficient calibration data is available. I tie these methods together using an estimation framework and effectively combine the calibration and reconstruction into a single optimization problem.
Next, I outline my current research exploiting sparsity to prospectively correct for head motion in functional MRI. Such motion affects the accuracy of time series correlations common in functional MRI analysis. I register frames as they are acquired and adjust the scan prescription prospectively for the detected motion. Through these examples, I demonstrate the great benefits model-based reconstruction holds for magnetic resonance imaging and other imaging modalities.
About the Speaker
Daniel Weller received his B.S. in Electrical and Computer Engineering with honors from Carnegie Mellon University in 2006, and his S.M. and Ph.D. in Electrical Engineering from MIT in 2008 and 2012. Daniel is currently a postdoctoral research fellow at the University of Michigan, supported by an NIH NRSA postdoctoral fellowship. He previously received a National Defense Science and Engineering Graduate (NDSEG) fellowship and an NSF graduate research fellowship. Daniel was a finalist in the student paper competition at the 2011 IEEE International Symposium on Biomedical Imaging. His research interests include magnetic resonance imaging, signal processing and estimation theory, nonideal sampling and reconstruction.
Architectures for Message-Passing Decoders
Speaker: Warren J. Gross, McGill University
Time: Monday April 8, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Error-correcting decoders based on message-passing algorithms over graphs are fundamental building blocks in modern communications and data storage systems. In this talk I will present recent results in developing architectures and implementations for high-throughput and low-power message-passing decoders. The talk will be divided into two parts: designing hardware for coding, and then applying coding to the design of hardware.
I will present four main architectural ideas. The first is stochastic decoding, where information is represented by the statistics of bit streams, resulting in simple, high-speed hardware implementations of graph-based decoding algorithms. Next, I will describe an architecture that allows multi-Gbps decoding of very long polar codes. I will then present an analysis of LDPC codes when the decoder is built exclusively out of faulty computing devices. and describe an extension to the Gallager-B algorithm that provides large gains in fault tolerance for a small decoding complexity overhead. Finally, I will show how concepts from message-passing decoding can be used to implement low-power network routers.
About the Speaker
Warren J. Gross received the B.A.Sc. degree in electrical engineering from the University of Waterloo, Waterloo, Ontario, Canada, in 1996, and the M.A.Sc. and Ph.D. degrees from the University of Toronto, Toronto, Ontario, Canada, in 1999 and 2003, respectively. Currently, he is an Associate Professor with the Department of Electrical and Computer Engineering, McGill University, Montreal, Quebec, Canada. His research interests are in the design and implementation of signal processing systems and custom computer architectures.
Dr. Gross is currently Vice-Chair of the IEEE Signal Processing Society Technical Committee on Design and Implementation of Signal Processing Systems and serves as Associate Editor for the IEEE Transactions on Signal Processing. He has served as Technical Program Co-Chair of the IEEE Workshop on Signal Processing Systems (SiPS 2012) and as Chair of the IEEE ICC 2012 Workshop on Emerging Data Storage Technologies. Dr. Gross is a Senior Member of the IEEE and a licensed Professional Engineer in the Province of Ontario.
Towards Clinically Viable Brain Machine Interfaces
Speaker: Vikash Gilja, Stanford University
Time: Tuesday April 9, 11:00 am
Location: Room 10.099, Two MetroTech Center, 10th Floor, Brooklyn
Abstract
Brain-machine interfaces (BMIs) translate neural activity into control signals for guiding prosthetic devices, such as computer cursors and robotic limbs, offering disabled patients greater interaction with the world. BMIs have recently demonstrated considerable promise in proof-of-concept animal experiments and in human clinical trials. However, a number of challenges for successful clinical translation remain, including system performance and robustness across time and behavioral contexts.
In this talk I will address these challenges by describing two classes of BMI experiments with rhesus monkeys. In these experiments we record from neurons in motor cortex using chronically implanted electrode arrays. The first class of experiments focus on control algorithm design. Through real-time closed-loop BMI experiments we demonstrate methods that increase performance and improve robustness. In the second class of experiments, we develop and verify a set of novel wireless neural recording systems, enabling the study of neural activity for longer time periods and across more complex behaviors. In addition to describing this work in animal model, I will introduce some of our recent work with human participants oriented towards the clinical translation of BMI.
About the Speaker
Vikash Gilja is currently a research associate in the Neural Prosthetics Translational Laboratory at Stanford University, working with Krishna Shenoy, Ph.D., and Jaimie Henderson, M.D. He received the B.S. degree in Brain and Cognitive Sciences and the B.S./M.Eng. degrees in Electrical Engineering and Computer Science from MIT in 2003 and 2004. In 2010, he completed his Ph.D. in Computer Science at Stanford University with the thesis "Towards Clinically Viable Neural Prosthetic Systems."
Controllable Magnetic Nanofluids and their Applications
Speaker: Shahriar Khushrushahi, Massachusetts Institute of Technology
Time: Wednesday April 10, 11:00 am
Location: Brooklyn
Abstract
Nanoparticles in a colloidal suspension do not fall out of solution due to two reasons. Their small size (≈10 nm) allows them to be easily dispersed by Brownian motion, counteracting gravity, and their surfactant layer (1-2 nm) prevents them from agglomerating. If the nanoparticles are magnetic, the resulting magnetic nanofluid exhibits a rich set of behavior that can be externally controlled by magnetic fields. The non-invasive nature of magnetic fields makes magnetic nanofluids ideal for several application areas, particularly in developing novel drug delivery devices. We also show that by using an oleophilic (oil-loving) surfactant layer, magnetic nanoparticles can also be used for energy and environmental applications such as cleaning and recovering oil from oil spills. Other interesting physical properties of magnetic nanofluids and their novel applications will also be presented.
About the Speaker
Dr. Shahriar Khushrushahi received his Ph.D. degree from the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology (MIT) in 2010, conducting research in the Laboratory of Electromagnetics and Electronics Systems (LEES). He also obtained his M.S. from MIT (2006) and his B.S. from Georgia Institute of Technology (2003), both in electrical engineering. He is currently a postdoctoral fellow in the Research Laboratory of Electronics (RLE) at MIT. His research interests include electromagnetics, ferrohydrodynamics, and magnetic nanofluid applications. His recent work has developed a novel magnetic separation system for cleaning up oil spills using magnetic nanofluids and permanent magnets. He recently received the Young Scientist Award at the 13th International Conference on Magnetic Fluids, and his oil spill recovery work has been showcased on CNN and the Discovery Channel.
Wireless Video: The Applications, the Challenges, and the Way Forward
Speaker: Professor Jerry D. Gibson, University of California, Santa Barbara
Time: Thursday April 11,
Location: LC400, Dibner Building, Five MetroTech Center, Brooklyn
Abstract
We examine the projected exponential growth in wireless video through 2017. It is suggested that streaming video is not the only application of interest and that perhaps the growth in wireless video is not as daunting as it appears at first blush. Advances in digital cellular standards are highlighted, but it is cautioned that when the Base Station (NodeB or eNodeB) is involved, expectations should be lowered. Three rules for improving wireless video performance and efficiency are illustrated with specific examples, and the next new challenge in wireless video is defined. The three principal take aways from the talk are that it is necessary to understand the full technology chain to provide a solution, more than one solution should be provided, and the solutions should not be too disruptive to gain traction.
About the Speaker
Jerry D. Gibson is Professor and Chair of Electrical and Computer Engineering at the University of California, Santa Barbara. He has been an Associate Editor of the IEEE Transactions on Communications and the IEEE Transactions on Information Theory. He was President of the IEEE Information Theory Society in 1996, and he has served on the Board of Governors of the IT Society and the Communications Society. He was a member of the Speech Technical Committee of the IEEE Signal Processing Society from 1992-1994, and he is currently a member of the Multimedia Communications and Wireless Technical Committees of the Communications Society. He was an IEEE Communications Society Distinguished Lecturer for 2007-2008, a member of the IEEE Awards Committee (2008-2010), and a member of the IEEE Medal of Honor Committee (2009-2010). He is an IEEE Fellow, and he has received The Fredrick Emmons Terman Award (1990), the 1993 IEEE Signal Processing Society Senior Paper Award, the 2009 IEEE Technical Committee on Wireless Communications Recognition Award, and the 2010 Best Paper Award from the IEEE Transactions on Multimedia.
Merging Basic Research in Data-Driven Cognitive Neuroscience with Real-life Application
Speaker: Jason Sherwin, Columbia University
Time: Friday April 12, 11:00 am
Location: JAB 473, Jacobs Academic Building, Six MetroTech Center, Brooklyn
Abstract
Finding the overlap between basic research and real-life application remains a challenge for any branch of engineering. But in the budding discipline of neural engineering, insights from basic research have ready use in as many disciplines as the human neural system is capable of operating. The challenge for us as engineers is to find those applications. In my talk, I will address how data-driven cognitive neuroscience can be used to gain insights into how the neural system functions in a few of such disciplines, namely music, combat and sports. Along the way, I will demonstrate the crucial role that machine learning and optimization play in creating increasingly robust decoding of neuroimaging data. Following the thesis that the neural system can be a guide to handling complex data in these and other environments, I will show how these results have ready application to the disciplines in which they were found and how they create a unique opportunity for both academia and industry to come together in a timely blend of i2e (invention, innovation and entrepreneurship).
About the Speaker
Jason Sherwin, Ph.D. holds dual appointments as a post-doctoral research scientist at the Columbia University in the City of New York and as an Oak Ridge Associated Universities post-doctoral fellow at the U.S. Army Research Laboratory (ARL). He also serves as the Managing Editor of the IEEE Transactions on Neural Systems and Rehabilitation Engineering. In addition to scholarly pursuits, he is active in the entrepreneurial community of New York, having served as consultant to Neuromatters, LLC and the City College of New York, each in their own respective contracts with the Defense Advanced Research Projects Agency (DARPA). His research covers perceptual decision-making in real-world environments, using en vivo neuroimaging and machine learning algorithms to improve the analysis of neural data in these complex and dynamic environments.
Application Oriented Networking (AON): Packet Processing with Multicore Processors
Speaker: Professor Laxmi N. Bhuyan, University of California at Riverside
Time: Friday April 19, 2:00 pm
Location: RH304, Rogers Hall, 6 MetroTech Center, Brooklyn
Abstract
Application Oriented Networking (AON) adds intelligence to routers and end points in the Internet by processing packets on the fly. It transforms the traditional network from pure packet-level routing to application-level processing by performing several customized computations at different nodes. The packet payload processing can be computationally expensive that adds delays, reduces throughput, consumes power, and degrades the quality of service (QoS) of the packet flow. Our research over the years has developed techniques to tackle the problems of packet processing by designing efficient scheduling techniques for multicore architectures. All our techniques have been implemented and tested on commercial multicore platforms. This presentation will introduce the problem and present comprehensive results obtained by us with Netbench, Webserver, Deep Packet Inspection (DPI) and multimedia applications. Our scheduling algorithm is based on Highest Random Weight (HRW), which maintains the connection locality for the incoming traffic, but only guarantees load balance at the connection level. We extended HRW to provide better load balancing and fairness, consider cache topology, and reduce the energy consumption.
About the Speaker
Laxmi Narayan Bhuyan is Distinguished Professor and Chairman of Computer Science and Engineering Department at the University of California, Riverside (UCR). Prior to joining UCR in January 2001, he was a professor of Computer Science at Texas A&M University (1989-2000) and Program Director of the Computer System Architecture Program at the National Science Foundation (1998-2000). He has also worked as a consultant to Intel and HP Labs.
Dr. Bhuyan served as the Editor-in-Chief of the IEEE Transactions on Parallel and Distributed Systems (TPDS) from 2006 to 2009. He is a past Editor of the IEEE TC, JPDC, and Parallel Computing Journal. Dr. Bhuyan is a Fellow of the IEEE, a Fellow of the ACM, a Fellow of the AAAS (American Association for the Advancement of Science), and a Fellow of the WIF (World Innovation Foundation). He has also been named as an ISI Highly Cited Researcher in Computer Science. He has received other awards such as Halliburton Professorship at Texas A&M University, and Senior Fellow of the Texas Engineering Experiment Station. He was also awarded the IEEE CS Outstanding Contribution Award in 1997. He was inducted into the Distinguished Alumni Hall of Fame of the Wayne State University College of Engineering in October 2010. He received the Distinguished Alumnus Award from National Institute of Technology, Rourkela in 2011.
Polar Codes and Pricing via Quantization
Speaker: Edmund Yeh, Northeastern University
Time: Monday April 22, 11:00am
Location: RH304, Rogers Hall, 6 MetroTech Center, Brooklyn
Abstract
Achieving the fundamental capacity limits of noisy communication channels with low complexity coding schemes has been a major challenge for over 60 years. Recently, a new coding construction, called polar coding, has been shown to provably achieve the capacity of discrete memoryless single-user channels. In the first part of the talk, we extend the polar coding method to two-user multiple-access communication channels. We show that if the two users use the channel combining and splitting construction, the resulting multiple-access channels will polarize to one of five possible extremals, on each of which uncoded transmission is optimal. Our coding technique can achieve some of the optimal transmission rate pairs obtained with uniformly distributed inputs. The encoding and decoding complexity of the code is O(n log n) with n being the block length, and the block error probability is roughly O(2^{-\sqrt{n}}). Our coding construction is one of the first low-complexity coding schemes which have been proved to achieve capacity in multi-user communication networks.
In the second part of the talk, we apply concepts from information theory to solve a canonical nonlinear pricing problem in microeconomics with information constraints. Here, a seller off ers a menu with a finite number n of choices to a continuum of buyers with a continuum of possible valuations. By revealing an underlying connection to quantization theory, we present the necessary conditions that the optimal finite menus for the socially eff icient and for the revenue-maximizing mechanism, respectively, must satisfy. In both cases, we show that the loss resulting from using the n-class finite menu converges to zero at a rate proportional to 1/n^2 as n becomes large. We then extend our nonlinear pricing model to the multi-product environment, where vector quantization can be used to jointly designing finite menus in multiple dimensions.
Join work with Eren Sasoglu (UCSD), Emre Telatar (EPFL), Dirk Bergemann (Yale), Yun Xu (Yale), and Ji Shen (London School of Economics)
About the Speaker
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 under Professor Robert Gallager in 2001. Since July 2011, he has been Associate Professor of Electrical and Computer Engineering at Northeastern University. Previously, he was Assistant and Associate Professor of Electrical Engineering, Computer Science, and Statistics at Yale University. He has held visiting positions at MIT, Princeton, University of California at Berkeley, Swiss Federal Institute of Technology Lausanne (EPFL), and Technical University of Munich.
Professor Yeh is the recipient of the Alexander von Humboldt Research Fellowship, the Army Research Office Young Investigator Award, the Winston Churchill Scholarship, the National Science Foundation and Office of Naval Research Graduate Fellowships, the Barry M. Goldwater Scholarship, the Frederick Emmons Terman Engineering Scholastic Award, and the President's Award for Academic Excellence (Stanford University). He is a Senior Member of the IEEE, a member of Phi Beta Kappa and Tau Beta Pi. He received the Best Paper Award at the IEEE International Conference on Ubiquitous and Future Networks (ICUFN), Phuket, Thailand, July 2012. Professor Yeh serves as the Secretary of the Board of Governors of the IEEE Information Theory Society.
A panorama on multiscale geometric representations, intertwining spatial, directional and frequency selectivity
Speaker: Laurent Duval, IFP Energies nouvelles, France
Time: Thursday May 23
Location: Brooklyn
Abstract
The quest for "optimal" representations in image processing and computer vision remains an very active area. The standard tasks of compression, denoising, restoration, require decompositions that trade off between efficiency and complexity, while achieving accurate rendering of smooth regions as well as reproducing faithful contours and textures. The inherent notion of scale in images has been satisfactorily captured by the large family of wavelets and pyramidal structures. Its most recent heirs (e.g. contourlets, curvelets, shearlets, dual-tree complex wavelets, etc.), born in the past 15 years, share a hybrid heritage highlighting the multiscale and oriented nature of edges and patterns in images. Such transforms typically exhibit redundancy, to improve sparsity in the transformed domain, and sometimes invariance with respect to simple geometric deformations (translation, rotation). This talk presents a panorama of these recent works on decompositions in multiscale, multi-orientation bases or dictionaries. [Joint work with Laurent Jacques, Caroline Chaux, Gabriel Peyré]
Additional information:
http://www.sciencedirect.com/science/article/pii/S0165168411001356
About the Speaker
About the Speaker: Laurent Duval is a research amateur at IFP Energies nouvelles and occasionally teaches signal and image processing. He received a State Engineering degree in electrical engineering from École supérieure d'électricité (Supélec), a master's degree in pure and applied mathematics from Paul Verlaine University (Metz, France), and the PhD degree in signal processing from the University of Paris-Sud on the topic of seismic data compression.
In 1998, he worked as a Research Assistant in the Multi-Dimensional Signal Processing Laboratory (MDSP Lab) at Boston University, Boston, MA. Since 2000, he conducts research in signal processing and image analysis with applications to geophysics, material characterization, analytical chemistry, and engine diagnosis. His research interests are in the area of non-stationnary digital signal and image processing, with a special emphasis on geometric wavelets, filter banks and time-frequency techniques, and their applications to denoising, filtering, detection, and data compression.