Department Seminars Spring 2008
Usually 12:30 p.m. Tues., EE125 |
Date & Speaker |
Topic |
| Jan 22: Vicky Mahodaya |
Augmentative Robotic Arm Control |
 |
| Feb 12 |
ECE Seminar |
| Feb 19: Alan R. Denton |
Colloidal Crystals: From Physical Properties to Engineering Applications |
|
| Mar. 13*: Sanghamitra Roy |
Numerically Convex Form |
|
| Mar. 13*: Koushik Chakraborty |
Over-provisioned Multicore Systems |
|
| Mar. 18*: Pradeep Kiran Sarvepalli |
Operator Quantum Error Correcting Codes |
|
| Apr. 1*: Dr. Cristinel Ababei |
Modern Physical Synthesis of VLSI Circuits and 3D FPGAs |
|
| Apr. 4*: Hongxiang Li |
Wireless Hybrid Network Design and OFDMA Capacity Analysis |
|
| Apr. 7*: Dr. Jingxian Wu |
Over-sampled Multi-carrier Modulation in Doubly Selective Fading |
|
| Apr. 11*: Dr. Samee Khan |
Optimizing the Energy Consumption and Performance of Computational Grids |
|
| Apr. 30*: Dr. Lingling Fan |
Control of Wind Turbine Generators |
|
| June 6*: Anupama Sadasiva |
An Analysis of the Role of Water on the Effective Permittivity of Biological Materials Using Mixing Formulas |
| Aug. 7*: Dr. Devendra Verma |
Design of Novel Nanocomposites for Bone Tissue Engineering |
| Aug. 8*: Dr. Xinnan Wang |
Micro/Nano Mechanical Characterization of One-dimensional Nanonaterials and Biomaterials |
| Aug. 18*: Dr. Yiquan Wu
|
Electrohydrodynamic Processing of Materials of Various Dimensions |
| Aug. 19*: Dr. Yechun Wang |
Biological Cells and Droplets in Microfluidic Sorting Devices and Vascular Vessels |
| Aug. 20*: Dr. Bonggeun Chung |
Microfabricated Platforms for Stem Cell Research |
| * Room and/or time different from the usual |
Augmentative Robotic Arm Control
12:30 p.m. Tuesday, Jan. 22, 2008
in EE 125
by
Vicky Mahodaya
NDSU
Fargo, ND
Abstract
This presentation will discuss completed and proposed work in the design and development of a smart robotic arm. A robotic arm has been instrumented to measure position, distance from an object, and gripping force on an object. Control algorithms will be developed so that the arm can
- detect an object within its working envelope,
- calculate the location of the object,
- shift from user-driven to algorithm-driven control to approach the object,
- grasp the object with a specified force,
- move to a predefined position, and
- return control back to the user.
Fuzzy logic control is proposed to control three axes of rotation and the grasping action to achieve the desired effects. An application of this work is the control of a prosthetic limb using a brain-computer interface.
Top
Colloidal Crystals: From Physical Properties to Engineering Applications
12:30 p.m. Tuesday, Feb. 19, 2008
in EE 125
by Alan R. Denton
Department of Physics NDSU
Fargo, ND
Abstract
Colloidal particles (nanometers-microns in size) suspended in a solvent by Brownian motion, when sufficiently concentrated, can self-assemble into periodic crystal structures. Examples include regular arrays of silica microspheres in natural opal gems and of polymeric particles in synthetic colloids. At the liquid-solid transition, thermodynamic stabilities of competing crystal symmetries depend sensitively on the nature of interparticle forces and external fields. Although mechanically fragile, colloidal crystals -- with lattice spacings comparable to visible-light wavelengths -- exhibit iridescence and other remarkable optical properties. Potential technologies include optical switching, photonic band-gap materials, chemical sensors, and electronic ink in high-resolution displays. After briefly introducing the physics of colloidal suspensions, I will focus on some practical engineering applications of charged and dipolar colloidal crystals.
Top
Numerically Convex Form: A new paradigm for optimizing discrete data with application to VLSI design automation
3:00 p.m. Thursday, Mar. 13, 2008
in Sudro 27
by Sanghamitra Roy
Department of ECE
University of Wisconsin
Madison
Abstract
Today's IC design involves dealing with complex VLSI systems and increasingly large number of design constraints, in a restricted design time for shorter time-to-market. Consequently, the design of high-performance low-power integrated circuits with millions of design components relies on efficient design automation techniques and algorithms. My research interests are focused on developing efficient algorithms and modeling techniques for applying mathematical optimization to nano-scale VLSI designs.
In this talk, I will propose a novel computer aided design flow using a new mathematical model "Numerically Convex Form", which advocates a fundamentally new paradigm for optimizing discrete data using geometric programming techniques. This model can be used in a wide realm of applications, which perform convex optimization with discrete data, within and outside the domain of VLSI design. I will describe an algorithm for modeling discrete data in this form with guaranteed minimum error, using the semidefinite programming framework. I will illustrate the application of the new mathematical model in one of the most critical VLSI design automation problems: the gate sizing problem for both combinational and synchronous sequential circuits. I will conclude the talk with some of my future research ideas.
Top
Over-provisioned Multicore Systems
4:00 p.m. Thursday, Mar. 13, 2008
in Sudro 27
by Koushik Chakraborty
Department of Computer Science
University of Wisconsin
Madison
Abstract
Technology scaling has provided system designers with an exploding transistor budget, far more than what was available when the core principles behind many existing commodity microprocessors were envisioned. An increasingly prominent trend that I observe indicates a substantial drop in the fraction of active chip area in future generation microprocessors. This trend will impact several fundamental principles of computer architecture, where architects must now consider novel use of inactive area for meeting their design goals.
In this talk, I will illustrate one embodiment of this new design paradigm in Over-provisioned Multicore Systems (OPMS). By design, OPMS provision more core resources than allowed by their power envelope. Computation is dynamically assigned to a subset of these processing cores, allowing individual cores to transition between active and inactive states. To demonstrate the potential of this design, I employ Computation Spreading that attempts to collocate similar computation fragments from multiple threads, while distributing dissimilar fragments from a single thread in a multithreaded application. Such a locality enhancement technique improves the performance and energy efficiency of the system by effectively using inactive core resources. In summary, OPMS designs enable a tremendous flexibility in assigning computation on processing cores, creating a framework for innovative techniques in future microprocessors.
Top
Operator Quantum Error Correcting Codes
9:30 a.m. Tuesday, Mar. 18, 2008
in Sudro 21
by
Pradeep Kiran Sarvepalli
Department of Computer Science
Texas A&M University
College Station
Abstract
Quantum computers are believed to give exponential speedups over classical computers for certain tasks such as integer factorization. However, the extreme sensitivity of quantum information has been a major obstacle to building a quantum computer. Fortunately, we can protect quantum information by the use of quantum error correcting codes. After a brief introduction to quantum coding theory, I will describe my work on operator quantum error correction. This is a recent development that generalizes both passive and active quantum error correction. It also promises to afford simpler error recovery schemes.
I will show how these codes, which were originally developed in the context of operator algebras, can be related to the familiar classical error correcting codes. I will show that it is possible to construct quantum codes from any arbitrary classical code unlike many previous constructions, which required self-orthogonal classical codes. I will conclude with a discussion on the relevance of these codes for fault tolerant quantum computation and directions for future research.
The talk does not presume any background in quantum computing.
Top
Modern Physical Synthesis of VLSI Circuits and 3D FPGAs
10:00 a.m. Tuesday, Apr. 1, 2008
in FLC 124
by
Dr. Cristinel Ababei
Design Engineer
Magma Design Automation
California
Abstract
In the first part of this talk, I will present CAD-level techniques that we used to address issues related to process variation. We investigated the relationship between robustness, predictability and performance of VLSI circuits at the placement level. Based on our analysis, we developed new methods for changing a standard timing-driven partitioning-based placement algorithm, which lead to designing more robust circuits with better predictability without sacrificing much of performance.
In the second part, I will discuss new FPGA architectures. More precisely, our objective is to investigate new 3D FPGA architectures for performance improvement, with emphasis on developing the necessary CAD tools. We designed and implemented a full flexible placement and routing tool for novel 3D FPGA architectures. The tool was used as a platform to exp27 January, 2009 11:41 AMechnologies can offer. Simulation experiments proved that 3D integration can significantly improve performance and total wire-length.
Brief Bio
Cristinel Ababei received his Ph.D. degree in electrical engineering from the University of Minnesota in 2004 and the B.S. degree in microelectronics from the Technical University of Iasi, Romania, in 1996. In 2004 he joined Magma Design Automation. His research interests include CAD for layout and logic synthesis for robust VLSI circuits and FPGAs, system-level design methodologies for systems-on-chip and embedded systems, and network optimization. He is a member of IEEE and ACM.
Top
Wireless Hybrid Network Design and OFDMA Capacity Analysis
3:00 p.m. Friday, Apr. 4, 2008
in FLC 124
by
Hongxiang Li
Electrical Engineering
University of Washington
Seattle
Abstract
Infrastructure based wireless networks can be conveniently categorized into two types. Broadcast: refers to the delivery of audio/video and other multimedia content over a reliable unidirectional channel (Example: DVB network); Unicast (one-to-one): refers to symmetric applications such as voice telephony and wireless data that require a bi-directional network and often has to contend with unreliable physical channels (Example: WiMAX and 3G-LTE). In the past, these two different classes of networks have evolved largely independently and are spectrum inefficient. This talk is about the imminent potential for convergence of these networks. Unlike cognitive radio (802.22) which provides non-collaborative opportunistic spectrum reuse for coexistence, I propose a joint design approach for broadcast and unicast to be simultaneously supported in a collaborative cellular network. In particular, I developed a dirty-paper coding (DPC) based overlay architecture to enable interference-free signal reception. The concept of multicarrier hybrid network is introduced to support both broadcast and unicast on a single-frequency platform. The fundamental capacity limits of hybrid communications are investigated, along with power allocation algorithms that approach the hybrid capacity. Our results show that OFDMA for unicast offers the best aggregate network capacity, and considerable capacity gains over uncoordinated multiplexing can be achieved through interference-free overlay. In addition, I provide approaches to further improve the spectral efficiency and the broadcast coverage through multi-cell collaboration. This work provides a clear pathway for achieving convergence of digital broadcasting and cellular data networks while delivering desired QoS objectives for multimedia distribution over a wide area.
Motivated by the increasing popularity of OFDMA, the second part of this talk studies the capacity optimality of OFDMA as a multicarrier multiple-access scheme. It addresses the relationship between OFDMA and the optimal multicarrier multiple-access schemes. In this talk, I will answer the following questions for both uplink and downlink SISO/ MIMO channels: (i) Is OFDMA optimal? If not, (ii) what are the conditions under which OFDMA is sum-rate optimal, and what are the probabilities of these conditions? (iii) in the case OFDMA is suboptimal, what is the performance gap between OFDMA and the optimal multicarrier multiple access solution? and (iv) Under the framework of OFDMA, how to allocate power and subcarriers to maximize the sum capacity?
Brief Bio
Hongxiang Li is a Ph.D. candidate in the department of Electrical Engineering at the University of Washington (UW), Seattle. He received the Bachelor of Engineering degree from Xi'an Jiaotong University, China in 2000, the M.S. degree from Ohio University, Athens OH in 2004, all in electrical engineering. At UW, his research focuses on wireless hybrid system design, network convergence, OFDMA/MIMO capacity analysis and cooperative communications. In summer 2006, he worked as a research intern in Radio Communication Lab at Intel, where he developed a novel interference detection algorithm for WLAN. Currently, he is doing an internship in the Technology Evolution and Strategy group at T-mobile USA.
Top
Over-sampled Multi-carrier Modulation in Doubly Selective Fading
3:00 p.m. Monday, Apr. 7, 2008
in FLC 124
by Dr. Jingxian Wu
Department of Engineering Science
Sonoma State University
Rohnert Park, CA
Abstract
With the ever increasing worldwide demand for broadband portable communications, wireless communication has enjoyed an explosive growth recently. The first part of this talk will give a brief overview about the challenges, opportunities, solutions, and plans for research and education in the area of wireless communications. Then, a new over-sampled multi-carrier modulation (OMCM) system operating in doubly selective (both time-selective and frequency-selective) fading is introduced as an example solution to many of the challenges faced by the design of the next generation wireless communication systems. The OMCM structure employs over-sampling in time domain and linear signal processing in frequency domain. The time-frequency processing enables a two dimensional Doppler-frequency grid, which allows the simultaneous achievement of time diversity and frequency diversity inherent in doubly selective fading. In addition, the new system has built-in mechanism that can suppress the inter-carrier interference (ICI) that has plagued multi-carrier communication systems in time-selective fading. Simulation results show that the new system outperforms conventional multi-carrier system with the same spectral efficiency by as much as 8 dB.
Brief Bio
Dr. Jingxian Wu is an Assistant Professor at the Department of Engineering Science, Sonoma State University, Rohnert Park, CA, USA. His research interests focus mainly on the physical layer of wireless communication systems, including multi-carrier communications, cooperative modulation, network information theory, space-time coding, channel estimation and equalization, and spread spectrum communications.
Dr. Wu is currently an Associate Editor for the IEEE Transactions on Vehicular Technology. Since 2006, he has served as a Technical Program Committee member for a number of international conferences, including the IEEE Global Telecommunications Conference 2006-2007, the IEEE Wireless Communications and Networking Conference 2007-2008, and the IEEE International Conference on Communications 2007-2009. He will serve as a symposium co-chair for IEEE Global Telecommunications Conference 2009.
Top
Optimizing the Energy Consumption and Performance of Computational Grids
3:00 p.m. Friday, Apr. 11, 2008
in FLC 124
by
Dr. Samee Khan
ECE Department
Colorado State University
Ft. Collins
Abstract
Energy consumption is a critical and crucial problem in large-scale computing systems, such as computational grids because they consume massive amounts of energy and have high cooling costs. These systems must be designed to meet functional and timing requirements while being energy-efficient. Resource allocation in computational grids is already a challenging problem due to the need to address deadline constraints and system heterogeneity. The problem becomes more challenging when energy management is an additional design objective because energy consumption of the system must be carefully balanced against other performance measures.
This talk focuses on the topic of resource allocation in computational grids with the aim to minimize energy consumption and makespan subject to the constraints of deadlines and architectural requirements. A solution from cooperative game theory based on the concept of Nash Bargaining Solution will be presented. In this game theoretical technique, machines collectively arrive at a decision that describes the best task allocation for the entire computational grid. This collective decision ensures that the allocations are both energy and makespan optimized. We also will look at some experimental results which verify that the cooperative game theoretical technique achieves superior performance compared to other traditional resource allocation techniques.
Brief Bio
Samee U. Khan is a postdoctoral research fellow in the Electrical and Computer Engineering Department at the Colorado State University. He received his Ph.D. degree in computer science from the University of Texas, Arlington in 2007.
His research interests include designing, building, analyzing, and measuring large-scale autonomous distributed computing systems using game theoretical and algorithmic mechanism design techniques, passive optical network layouts, designing secure systems, combinatorial games, and combinatorial optimization. His research work in these areas is published in 40 technical papers.
Dr. Khan is a member of the European Association of Theoretical Computer Science, the Game Theory Society, the IEEE Communications Society, the IEEE Computer Society, and the Society of Photo-Optical Instrumentation Engineers.
Top
Control of Wind Turbine Generators
1:00 p.m. Wednesday, Apr. 30, 2008
in CIE 205
by Dr. Lingling Fan
ECE Dept.
NDSU
Fargo
Abstract
Wind turbine generation becomes a booming technology after the milestone achievements in power electronics, namely, pulse width modulation and Digital Signal Processing control. Dr. Lingling Fan will present the wind turbine generator control technology in this seminar.
The objective of control is to extract maximum power from wind and to keep a constant voltage at the generator terminal. The talk will address choosing which type of generators, how to achieve control via power electronics and the art of controller design. Also in the talk, a brief look is given at power engineering as the multi-disciplinary area embracing operation research, control and electronics applications. Top
Design of Novel Nanocomposites for Bone Tissue Engineering
2:00 p.m. Thursday, Aug. 7, 2008
in Dolve 204
by Dr. Devendra Verma
Department of Civil Engineering
NDSU
Fargo
Abstract
More than 400,000 people undergo hip or knee replacement surgeries every year in the USA alone. This number is anticipated to go even higher in coming years. Acute shortage of transplantable donor tissue and limited life span of bone implants (10-15 years) have prompted surgeons and scientists to look for viable alternatives. Tissue engineering has shown significant potential for regeneration of tissue or organs. Tissue engineering applies principles of engineering, biology and chemistry to regenerate a tissue. For tissue regeneration, cells are seeded in a highly porous three dimensional construct called scaffold. Biocompatibility, biodegradability, and ability to attach and promote cell growth are essential requirements of a scaffold. Specifically for bone tissue engineering, adequate mechanical response of scaffold is critical. Various polymers both of synthetic and biological origin have been investigated as a material for scaffold. Synthetic polymers are biodegradable, biocompatible, and can easily be formed into different shapes and sizes. However, hydrophobicity, lack of functional groups and release of acidic products on degradation are causes of concern. Biopolymers such as collagen, chitin, chitosan etc. promote cell adhesion, proliferation and differentiation, and evoke minimal foreign body reaction on implantation. But, they have inadequate mechanical properties for bone regeneration and tend to loose structural integrity under wet and body fluid conditions. The quest for scaffold materials which can, not only promote cell adhesion, proliferation and differentiation but also have adequate mechanical strength to support bone tissue growth is still on. We have designed novel nanocomposites with enhanced mechanical response for bone tissue engineering by tailoring interfaces. This talk will cover design principles and mechanism involved in enhancement of mechanical response. It will also cover cellular response studies done on these nanocomposites.
Top
Micro/Nano Mechanical Characterization of One-dimensional Nanonaterials and Biomaterials
2:30 p.m. Friday, Aug. 8, 2008
in Dolve 204
by Dr. Xinnan Wang
Department of Mechanical Engineering
University of South Carolina
Columbia
Abstract
This presentation focuses on the experimental micro/nano mechanics of two material systems: one-dimensional (1D) nanomaterials and biomaterials. The extremely small dimensions of 1D nanostructures and the soft nature of biological materials impose tremendous challenges to many existing testing and measuring techniques for experimental studies of their mechanical properties.
In part one, several powerful and unique testing methodologies are implemented on the representative 1D nanostructures. The apparent elastic moduli of the nanostructures are found less than that of their bulk entities. The physical mechanisms that underpin this variation are investigated and addressed.
Part two is centered on the micro/nano studies of deformation behaviors of biomaterials. Type-I collagen and virus-templated core-shell composites are two examples for probing the mechanical properties of interest. The deformation behaviors and the corresponding underlying mechanisms are discussed. The findings shown in the presentation are significant for providing the guidelines of further rational design of MEMS/NEMS and biomedical devices. Top
Electrohydrodynamic Processing of Materials of Various Dimensions
2:00 p.m. Monday, Aug. 18, 2008
in Dolve 204
by Dr. Yiquan Wu
Department of Mechanical Engineering & Materials Science
Duke University
Durham, NC
Abstract
The preparation of materials at the micro- and nanometer scale using innovative processing techniques is attracting research interest both in fundamental and applied sciences worldwide. Recently, electrohydrodynamic atomization processing has been used to prepare several types of micro- and nanometer materials, including particles, fibers, films, and 3D components. The structure of these materials can be controlled in ways that have great potential for biomedical, environmental, electronic, and energy-related applications.
There are two key types of electrohydrodynamic atomization processes: electrospraying and electrospinning. Electrospinning has been an effective and attractive processing technique in the generation of one-dimensional materials, and has a variety of applications in bioengineering and nanotechnology. Electrospun fibers at the micro- and nanometer scale can serve as templates for nanotubes, conductive wires, nanostructured electrolytes, gas storage materials, filters, and components in photo-catalysis, sensor, and electronic devices.
Electrospraying has the ability to generate mono-dispersed particles whose size may vary from tens of nanometers to hundreds of micrometers. The production of particles in the micro- and nano-scale with narrow size distributions and controlled structure are of interest for use as drug delivery carriers and in film preparation.
One of the great challenges of nanotechnology is the development of novel and cost-effective techniques for fabricating nano-scale components. Electrohydrodynamic atomization processing is believed to be a promising approach to address this challenge and help make the utilization of these exciting materials more accessible.
Top
Biological Cells and Droplets in Microfluidic Sorting Devices and Vascular Vessels
2:00 p.m. Tuesday, Aug. 19, 2008
in Dolve 204
by Dr. Yechun Wang
Department of Coatings and Polymeric Materials
NDSU
Fargo
Abstract
Biological cells have to be manipulated and characterized individually in studies of immunology, cell biology, molecular biology, and their clinical applications in AIDS, leukemia/lymphoma, and cancer detection. In the design of new enzymes, the in vitro transcription and translation of single genes are carried out via emulsions, or droplets, the size of which has to be precisely controlled. Microfluidic sorting devices have thus been widely employed to perform the afore-mentioned tasks.
To provide guidance in the design of simpler, faster and cheaper microfluidic sorting devices, the cell/droplet behavior is investigated in microfluidic channels under the influence of pressure-driven flow. A numerical model predicting the deformation, translation and migration of droplets in microfluidic channels will be presented. A novel three-dimensional spectral boundary element algorithm has been developed for the droplet motion in microfluidic devices. To achieve better stability during the droplet deformation, a suitable interfacial smoothing is developed based on a Hermitian-like interpolation. An adaptive mesh reconstructing procedure is also employed based on the relevant lengths of the spectral elements. The algorithm is validated by comparing the numerical results with experimental findings and analytical predictions. The droplet shape, velocity and lateral migration are predicted under the influence of flow velocity, interfacial tension and material properties.
In addition, the study on the hemodynamic forces exerted on endothelial cells or white blood cells attached to the surface of vascular vessels will be presented. This study investigates the relative importance and the nature of two components of the hemodynamic force, i.e., the shear and normal force, on the cell. We consider a wide range of vascular vessels (from capillaries to arteries) and the spreading angles of the cell. Based on computational investigation and scaling analysis, the study demonstrates that the normal force contributes significantly to the total force on the cell and its influence is much more substantial in small vessels. The results may also be applied to the fluid forces on protuberances in microfluidic devices. Top
Microfabricated Platforms for Stem Cell Research
2:00 p.m. Wednesday, Aug. 20, 2008
in Dolve 204
by Dr. Bonggeun Chung
Harvard-MIT Division of Health Sciences and Technology
Cambridge, MA
Fargo
Abstract
This presentation describes the development and characterization of microfluidic platforms to study proliferation, differentiation, migration, and apoptosis of neural stem cells (NSCs). NSCs hold tremendous promise for fundamental biological studies and cell-based therapies in human disorders. NSCs are defined as cells that can self-renew yet maintain the ability to generate the three principal cell types of the central nervous system such as neurons, astrocytes, and oligodendrocytes. Despite their promise, cell-based therapies are limited by the inability to precisely control their behavior in culture. Compared to traditional culture tools, microfluidic platforms can provide much greater control over cell microenvironments and optimize proliferation and differentiation conditions of cells exposed to combinatorial mixtures of growth factors. NSCs proliferated and differentiated in a graded and proportional fashion that varied directly with growth factor concentration. Directed embryonic stem (ES) cell differentiation is also a potentially powerful approach for generating a renewable source of cells for regenerative medicine. Typical in vitro ES cell differentiation protocols involve the formation of ES cell aggregate intermediates called embryoid bodies (EBs).
Recently, we demonstrated the use of poly(ethylene glycol) (PEG) microwells as templates for directing the formation of these aggregates, offering control over parameters such as size, shape, and homogeneity. Despite these promising results, the previously developed technology was limited as it was difficult to reproducibly obtain cultures of homogeneous EBs with high efficiency and retrievability. In this study, we improve the platform by optimizing a number of features: material composition of the microwells, cell seeding procedures, and aggregate retrieval methods. Adopting these modifications, we demonstrate an improved degree of homogeneity of the resulting aggregate populations and establish a robust protocol for eliciting high EB formation efficiencies. Therefore, the development of microfluidic platforms and hydrogel microwell arrays will help in advancing our understanding of brain development and provide a versatile tool with basic and applied studies in stem cell biology. Top
|
Welcome