http://ilp-www.mit.edu/ List of 40 most recent project additions, as compiled by the Industrial Liaison Office. en-us Copyright 2009 MIT ILP Sat, 7 Nov 2009 09:15:32 GMT http://ilp-www.mit.edu/favicon.ico http://ilp-www.mit.edu/display_page.a4d?key=P5d http://ilp-www.mit.edu/display_project.a4d?projectId=20506 Join the Prajnopaya Foundation to do work in one of the following areas: (*) Tuberculosis Treatment and Prevention Project (*) Tsunami Rehabilitation (* )Upaya Care The Pranjopaya Foundation is a worldwide humanitarian organization, with the purpose of alleviating suffering and caring for society by developing innovative health, educaiton and social welfare projects. 11/04/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20505 In June, 2009, the International Iberian Nanotechnology Laboratory (INL) and MIT began a major new collaboration to enrich each institution’s research activities in nanoscience and nanotechnology. MIT-INL, a new education and research enterprise focusing on nanotechnology, will enable approximately $35 million (25 million euro) of new sponsored research with MIT in its first five years. The MIT-INL agreement leverages the Institute’s especially strong reputation in materials science, engineering, nanotechnology and biotechnology. Active research projects include work on nanoparticles that can selectively adsorb water contaminants, autonomous microsystems that can move around water supplies and sense contaminants (while sustaining themselves on power scavenged from their environments), new materials for energy storage, revolutionary tools and technologies for monitoring our food supply, and others. Students can apply for summer or IAP research interships through MIT-INL, and there will be periodic Institute-wide calls for proposals from faculty for new research projects. INL is an international research facility located in Braga, Portugal, and is a joint project of the governments of Portugal and Spain. allows for the exchange of students and faculty 11/04/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20503 The MIT-Brazil Program launched in the fall os 2009 as the tenth MISTI country program. MIT-Brazil is open to all MIT students -- all majors; undergraduate and graduate students. 10/30/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20502 The Working Group on the Strategic Use of Force seeks to convene individuals interested in studying both state and sub-state actors' strategies and tactics to deploy force or coercion in order to achieve their objectives. This group is interested both in the strategies themselves and a variety of conditions that influence the strategic effectiveness, including civil-military relations, alliances, and regime type. This is an MIT working group for individuals with research interests in violent state and non-state actors and will serve as a forum for students, faculty, and outside researchers to present their ongoing work and received feedback. Its aim is to examine why particular actors select the coercive strategies they do, as well as the means and tactics they use. Topics may include but are not limited to the causes, organization and consequences of inter-state conflict, inter-state coercion, deterrence, military intervention, as well as studies of sub-state violence including civil war, terrorism, and insurgency. The working group is meant to cover a number of questions and theories on security within the fields of international relations and comparative politics that specifically deal with the threat or application of force. 10/29/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20501 The MIT-Africa Program was established with a gift of Total’s gift of a Chair on Contemporary Africa to the School of Humanities and Social Sciences as a first step in a series of collaborations in education and research between Africa and MIT. Professor Suzanne Berger, MISTI Director, and Patricia Gercik, MISTI Associate Director, are currently managing the Program. In Fall 2008, MISTI provided partial support for Sloan School Dr.Anjali Sastry's course "Practicing Management" -- a section of the Locke/Johnson G-Lab class that focuses on Health Distribution in Africa. Sixty students took the class last fall and then spent January in hands-on projects in African hospitals and clinics during January 2009 IAP. MISTI has provided some support for Africa Information Technology Initiative (AITI), a student-run organization that promotes development in Africa through education in appropriate information and communication technologies (ICTs). During MIT's summer recess, AITI sends MIT students to Africa to teach African undergraduate and high school students. AITI partners with local African institutions to offer classes focused on mobile phone application development with an emphasis on independent research, problem-solving, and entrepreneurship. In addition, MISTI has provided funding for supported the iLabs Project in Africa. 10/29/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20499 10/29/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20498 The Global Sustainability Working Group (GSWG) brings together graduate students, faculty and research scientists from across MIT to discuss global sustainability challenges and appropriate governance frameworks. The group is a continuation of the Environmental Vulnerability, Resilience, and Justice group founded in 2008. The name has been changed to reflect the genesis that has taken place in the scope of inquiry. The GSWG looks across a variety of scales and disciplines and from diverse methodological and epistemological standpoints to tackle questions of environmental, social and economic sustainability. We consider existing and emerging theoretical advancements as well as how researchers and practitioners can contribute to the creation of more sustainable cities and regions in both the global North and South. In 2009-10, this group will meet on the last Tuesday of each month during the academic year, from 5.30 pm to 7.30 pm at the Center for International Studies. The sessions will continue to be anchored in discussions of student and faculty work in progress. The intent is to encourage analytic discussions and critical feedback among the members of the group and invited participants. Work in progress encompasses new faculty projects, doctoral and masters-level student papers, research reports, and dissertation and thesis projects at different stages of development. Once each semester, we will invite faculty members and researchers from the Boston area whose work is of particular relevance to the global environmental sustainability topic so that we can explore emerging areas of research of interest to members of the working group. 10/29/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20497 MIT and National Research Foundation (NSF) of Singapore today announced the launch of a project to develop new models and tools for the planning, design, and operation of future urban transportation. Aimed at making urban transportation systems more environmentally sustainable -- first in Singapore, and ultimately on a global scale -- these new models will be developed and deployed by nearly 60 researchers from four academic institutions. The five-year project will be led by Amedeo Odoni, Professor of Aeronautics and Astronautics and of Civil and Environmental Engineering, and engage some 30 other faculty and researchers from the School of Engineering, the Sloan School of Management, and the School of Architecture and Planning at MIT. Assisting their efforts will be approximately 25 faculty members from the National University of Singapore, Nanyang Technological University, and the Singapore Management University. This project will be a significant increase in the scale of transportation-related research conducted by MIT faculty and students. At the heart of the Singapore project is SimMobility, a simulation platform with an integrated model of human and commercial activities, land use, transportation, environmental impacts, and energy use. This simulation will be linked with a range of networked computing and control technology-enabled mobility innovations. The project's researchers plan to use the data generated by these devices, and a range of new analytical tools that harness real-time information and management systems, to design and evaluate new mobility solutions for urban settings in and beyond Singapore. The Future of Urban Mobility team is the fourth interdisciplinary research group in the Singapore-MIT Alliance for Research and Technology Centre, or SMART Centre. The first three groups are in biosystems and micromechanics, environmental sensing and modeling, and infectious diseases. SMART is MIT's largest international research endeavor and the first research center of its kind located outside Cambridge, Massachusetts. 10/28/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20496 An all liquid metal grid-scale battery for low cost, large scale storage of electrical energy. This new class of batteries could enable continuous power supply from renewable energy sources, such as wind and solar and a more stable, reliable grid. 10/27/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20494 MIT has been awarded a five-year grant from the National Cancer Institute (NCI) to start a new Physical Science-Oncology Center. The funding, approximately $3.5 million per year, will support four cancer research projects led by MIT physical scientists. The NCI announced yesterday that 12 institutions, including MIT, will host the new centers. The Physical Science-Oncology Centers will take new, non-traditional approaches to cancer research by studying the physical laws and principles of cancer, and applying those principles to understanding the evolution of cancer and unraveling cancer's complexity. MIT's new center will also include investigators from the Whitehead and Broad institutes, Harvard Medical School, Brigham and Women's Hospital and Boston University, as well as the University of California at San Francisco, Stanford, and the Hubrecht Institute in the Netherlands. The projects bring an array of physical sciences techniques to bear on cancer research, including computational modeling, statistical analysis, and novel sensors that can track the activities of single cells. Van Oudenaarden will be working with Tyler Jacks, director of the David H. Koch Institute for Integrative Cancer Research at MIT, to study how colon stem cells become cancerous. Van Oudenaarden has already developed a technique to follow single cells and observe how their gene expression changes over time. The grant will also fund a project by associate professor of biological engineering Scott Manalis, who is collaborating with Harvard Medical School professor Marc Kirschner and with Koch Institute professor Angelika Amon to measure single cells over time. Manalis has built a sensor that weighs cells with unprecedented accuracy, which could allow the team to figure out the growth patterns of normal and cancerous cells. Arup Chakraborty, professor of chemical engineering, chemistry and biological engineering, will collaborate with cancer biologists Jeroen Roose, Kevin Shannon and Benjamin Braun of the University of California at San Francisco Medical School, to study how signaling networks in T cell lymphomas are altered by cancer-promoting mutations that result in overactivation of Ras proteins. They will use computational methods rooted in statistical physics to complement in vitro and in vivo experimentation. Leonid Mirny, associate professor in the Harvard-MIT Division of Health Sciences and Technology, will collaborate with Shamil Sunyaev of Brigham and Women's Hospital to study cancer as a micro-evolutionary process that takes place inside an organism. This project aims to test a novel approach for cancer treatment using cancer's intrinsic weakness -- mutations accumulated during the micro-evolutionary process -- to bring the population of cancer cells down. Key to the projects' success is the inclusion of traditional cancer biologists, who will help guide the physicists in their new approaches. 10/27/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20493 The physics of microscopic spin ensembles can be distinctly different from that of macroscopic ensembles. For example, in volumes of nuclear spins smaller than about < (100 nm)3, random spin flips generate a fluctuating polarization that exceeds the typical thermal (or Boltzmann) polarization. These spin fluctuations are a major source of dephasing in solid-state quantum systems, and their control is an important prerequisite for nanometer-scale magnetic resonance imaging (MRI) and spectroscopy. In our lab we focus on the fundamental physics that goes into these experiments on small numbers of spins. We also try to understand “back-action” effects from the detector on dynamics of spins as they become strongly coupled to, for example, a nanomechanical resonator. 10/26/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20492 Local chemical characterization of nanostructured surfaces, especially when working with organic materials like self-assembled monolayers or thin polymer films, is currently very hard or not possible. Due to the chemical specificity of magnetic resonance, it may be possible to apply spectroscopy techniques (like NMR) for exactly that purpose. The lab’s objective is to adopt such techniques so as to simultaneously image a surface's chemical composition (using spectroscopy), spin density (using Nano-MRI) and topography (using the sensor in force microscopy mode) with nanometer resolution. This application is also very interesting for semiconductor devices as well as for a variety of nanostructured surfaces in material and energy science, like catalysts. 10/26/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20491 Sensitive detection of magnetic forces requires mechanical structures that can measure very small forces. Some of the best sensitivities are achieved with micromechanical cantilevers, which reach detection limits in the range of Attonewtons (10e-18N/rtHz). Making even more sensitive mechanical structures is important not only for nanoscale spin detection, but also for mass sensing or testing of fundamental laws in e.g. gravity. The lab explores limitations to these detectors. This includes studies of the various dissipation mechanisms. We are also interested in resonators made from new materials. These might not only offer better force sensitivity, but by studying their behavior we also learn more about the material’s fundamental properties. 10/26/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20490 One of the main goals is the application of nanoscale spin detection to the 3D imaging of important biological structures by MRI. The main obstacles to these experiments are sensitivity and spatial resolution. Nanomechanical force sensors have, in the latest demonstration experiments, pushed the resolution of MRI to about 5 nm and enabled three-dimensional imaging of individual virus particles (a roughly 1000x improvement compared to conventional MRI techniques). The objective is to make these methods more widely applicable and to apply them to 1-100 nm-sized biological objects. Many of these objects are of extraordinary relevance to biology, including, for example, single virus particles, functional cellular units (like the ribosome), or Amyloid fibrils (implicated in Alzheimer’s disease). The aim is to use Nano-MRI for resolving their superstructure and eventually locating specific functional units using isotopic labeling techniques. Another strategy is the combination with atomistic data from X-ray analysis (that is generally incomplete for large molecules), with the aim of reconstructing full atomic structures. 10/26/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20489 Diamond crystals host a number of fluorescent defects. One of these defects, the so-called nitrogen-vacancy (NV) color center, has a single electronic spin associated with it. This spin can be controlled individually based on optically detected magnetic resonance, can be implemented in nanodiamonds less than ten nanometers in size, and can thus be used as sensitive magnetic detectors with nanometer spatial resolution. In our lab, we intend to integrate high-quality NV centers into scanning probes, so that we can apply them for directly mapping electron and nuclear spins on surfaces with nanometer resolution. Combination with magnetic resonance spectroscopy may then permit full chemical analysis of materials and study of surface dynamics. This "scanning diamond magnetometry" was proposed only very recently and is a genuine idea from the Degen Lab. 10/26/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20488 At the Degen lab we invent, build and operate tools for detecting weak magnetic signals at very small lengthscales. Our goal is to apply these tools for the following purposes: (*) Study of nanoscale spin physics, quantum effects, and other fundamental physical phenomena that become important at these small scales (in condensed matter and quantum physics) (*) Nanoscale MRI imaging of single biological objects 1-100 nm in size (in structural biology) (*) Chemical analysis of heterogeneous surfaces with nanometer resolution using spectroscopy (in surface science) Magnetic signals from spins in nanometer-sized samples are typically extremely small. The lab currently specializes on two experimental approaches that have the necessary sensitivity to examine these signals. In Magnetic Resonance Force Microscopy (MRFM), we take advantage of ultrasensitive cantilevers such as used in force microscopy to measure spins by their magnetic force. In recently invented Scanning Diamond Magnetometry, we rely on the exquiste magneto-optical properties of the NV defect in diamond to sense the same minute magnetic fields optically and under ambient conditions. Along the way of improving the sensitivity of our instrumentation, we also hope to make exciting new physical discoveries. 10/26/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20487 In collaboration with the lab of Job Dekker (UMass Medical School at Worchester) we examined architecture of DNA inside the cell. Recently developed method (Hi-C) probes 3D architecture of whole genomes by coupling proximity-based ligation with massively parallel sequencing. This method detects pairwise contacts between distant genomics regions. Our analysis of Hi-C data revealed that at the ~5 Mb scale, chromatin conformation is consistent with the unusual, fractal, non-equilibrium structure: crumpled (fractal) globule, introduced in polymer physics two decades ago. The fractal globule is defined as an unknotted compact polymer structure and is distinct from the classical equilibrium globule state of a polymer that is highly knotted. Using simulations we demonstrated that fractal globule is an attractive model for DNA organization. This knot-free conformation enables maximally dense packing while preserving the ability to easily fold and unfold any genomic locus. Moreover it provides “territorial” DNA organization such that regions close in the genome occupy the same volume in 3D. 10/22/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20486 As a part of a recently formed, NCI-funded center for Physics-in-Oncology, we developed a program in evolutionary dynamics of cancer. By considering neoplastic progression as a micro-evolutionary process we study accumulation of driver and passenger mutations in cancer. While the former are responsible for cancer progression and are in the spotlight of studies in oncology, the later may confer deleterious to cancer cells -- a vulnerability that can be exploited by therapeutics. By combining simulations of cancer progression with analysis of cancer genomic data we developed methods to characterize patterns of mutations in cancer, allowing to differentiate tumor suppressors from oncogenes, and passengers from drivers. We measured fitness effect of different classes of mutations demonstrating that passengers are subject of weakened purifying selection and are likely deleterious to cancer cells. The goal is to develop a theory of neoplastic evolution informed by cancer genomic and experimental data; use it as a framework for characterization of driver and passenger mutations by original statistical techniques, and to test feasibility of pushing a cancer into a population meltdown due to elevated mutation load. 10/22/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20485 Cell shape, motility and division in eukaryotes rely on the self-assembly of microtubules, long dynamic filaments. One of the most striking properties of microtubules is dynamic instability: their ability to switch between periods of growth and shrinkage. Several anti-cancer drugs work by interfering with this process and thus arresting cell division. The biophysical model of the process connects microscopic properties of tubulin dimers with macroscopic dynamics of the microtubule. By taking into account both mechanical properties of tubulin and the stochastic nature of polymerization reactions, we obtained a detailed picture of dynamic instability, providing new insights into the underlying physical mechanisms. In collaboration with the lab of Professor Mitchison and Professor Needleman (Harvard University), we developed a technique to infer some parameters of the dynamic instability process in vivo. The lab has developed a theoretical formalism that connected microscopic rates of the process with macroscopic observables that are measured by single-molecule fluorescent microscopy. The method allowed examining dynamics of microtubules in mitotic spindle, testing and quantitatively comparing a number of models of mitotic spindle assembly. 10/22/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20484 These include detection of knots in protein structures (jointly with Prof.Kardar); physical characterization of viruses and their capsids; further development of an algorithm to identify key residues in proteins using evolutionary information; development of web-servers to make these tools widely accessible; biophysical modeling of self-assembly of microtubules and development of the new experimental technique to monitor this process in vivo; and development of new algorithms for analysis of gene expression data. 10/22/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20483 We examined the structure of the first networks of protein-protein interactions, demonstrated their modular organization, and developed algorithms and statistical tools to detect such modules. We studied the structure and evolution of the metabolic network and demonstrated that the modules that evolve as a unit differ from traditional definitions of distinct metabolic pathways. We also developed a novel algorithm to study networks of chemical reactions and demonstrated that the topology of the metabolic network is sufficient to predict its response to perturbations. 10/22/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20482 The MIT Ideation Lab is a part of the Department of Mechanical Engineering and the Engineering Systems Division at MIT. Research focuses on design through several perspectives: (*) People: We consider the peoples and teams who conduct design, and the end users of designed artifacts (*) Products: Research investigates the assessment of design outcomes, both during their creation and after they are in use (*) Processes: Work examines the activities involved in designing as well as more formal methodologies for synthesizing design 10/20/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20481 In recent high performance processor design, long distance interconnects became a serious bottleneck under tight power constraint. Equalized on-chip interconnects have been presented significant improvement in performance without sacrificing too much power consumption. This work presents further improvement of energy efficiency of equalized-interconnect by proposing two circuit techniques: (1) pre-distorted charge-injection (CI) feed-forward equalization (FFE); 2) trans-impedance (TIA) termination at receiver. Instead of using traditional analog subtraction, CI-FFE injects pre-computed the current value required for FFE into the channel while mitigating the nonlinearity of the driver. The non-linearity of the driver is statically compensated by pre-distorting FFE coefficients. The trans-impedance amplifier terminated at the receiver improves the bandwidth, signal amplitude, and reduces bias current. A test-chip is fabricated in 90nm CMOS process and consumed about 0.4pJ/b running at 4Gb/s with vertical eye opening about 100mV and horizontal eye opening 50%UI differential peak-to-peak. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20479 Based on voltage-to-time conversion technique a pseudo-differential two-way-interleaved adaptive linear receive equalizer with two 2x-oversampled feed-forward taps has been designed in a 90-nm CMOS process. It integrates equalization and phase interpolation functions into one unit to simultaneously address inter-symbol-interference (ISI) cancellation and phase synchronization in a link receiver. Due to the process speed limitation, the half-rate time interleaving technique is also applied. Four sampling phases with 25% duty cycle are generated locally from another pair of quadrature clocks. A voltage-to-time (V2T) block converts the sampled signal into a delayed digital signal, transferring the sampled information into the time-domain. All four V2T converters are followed by a time-to-voltage (T2V) stage to realize summing, subtraction and multiplication. Equalizer tap weights are implemented as two programmable reference currents I1 and I2 biasing T2V blocks. Two slicers with tunable thresholds are added to sense the signs of the input signal and output error of the FSE, respectively, and to enable tap weight adaptation with external adaptive engine. The design is fabricated in a 90-nm CMOS process. It operates at 4 Gbps with 8 mW power consumption and linearity of 4.3 effective bits at 1.2 V supply. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20478 With the escalating costs of hospital visits, clinicians are opting to use at-home monitoring devices to diagnose patients. Current ECG Holter monitoring devices typically have 24 - 48 hour memory and battery capacity. With many patients experiencing intermittent heart problems that can occur once every week -- month, the Holter monitor is not a good solution and an event recorder or loop recorder is required. However, each of these recorders can only save up to a few minutes of ECG recordings. This leads to the loss of most of the data, which could be very important in alerting the user for the onset of future episodes. Therefore, we have developed a Holter monitor prototype with the goal of battery and memory capacity of two weeks. We based the long-term monitor prototype around a Texas Instruments MSP430 low-power microcontroller which enables high computing power with very low power consumption. The prototype monitor, which is currently being designed, will be mounted on standard 3M 2560 Red Dot electrodes and fabricated on a flexible PCB substrate. Mounting the PCB directly to the electrodes will improve the SNR by an estimated 40 dB compared to using wired leads. The monitor will be ‘L’ shaped with rounded corners and placed on the patient’s chest. The ‘L’ shape will enable several mounting sites to be placed on the board which will allow the doctor to choose which measurement he would like to record. The monitor will have 320 Mbytes of FLASH memory which is enough to store two weeks of data sampled at 250 Hz continuously. Total power consumption of the system is estimated to be less than 8mW. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20477 The area of Millimeter-wave (MMW) system research and design has become increasingly popular in recent years, as advanced silicon processes have enabled integrated circuit operation in the MMW regime. Several applications exist for MMW design, including wireless communications at 60-GHz, collision-avoidance radar imaging at 77-GHz, and concealed-weapons detection imaging at 94-GHz and higher. This research focuses on a passive imager front-end that has been developed and tested for the application of concealed-weapons detection. A key component of this research involves the design of a packaged antenna. A wideband Vivaldi-type design is used to achieve high gain and efficiency from 73-GHz to 105-GHz. The antenna was fabricated on a low-conductive dielectric constant material, RO4350B, at the MIT EML Laboratory. It is packaged onto the LNA bondpad terminals via gold solder bumps and silver epoxy and placed to maximize the distance from the die’s ground plane. The antenna has a gain of approximately 8dB within the operating bandwidth and a typical efficiency of 80%. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20475 The organic thin-film transistor (OTFT) is a field-effect transistor technology that uses organic materials as the semiconductor. OTFTs have field-effect mobilities that are comparable to those of hydrogenated amorphous silicon TFTs, and OTFTs are compatible with large-area and mechanically-flexible substrate. The goal of this work is to demonstrate an integrated OTFT temperature-sensing circuit suitable for large-area and flexible substrates. As shown, two important differences are observed between the OTFT’s and the MOSFET’s current-voltage characteristics when temperature is varied. First, the OTFT’s current increases with temperature in both subthreshold and above-threshold regimes, whereas the MOSFET’s above-threshold current decreases with temperature. Second, the OTFT’s subthreshold slope is temperature independent over the measured range of -20 to 60°C, while the MOSFET’s subthreshold slope is proportional-to-absolute-temperature (PTAT). Because of these differences in temperature response, the OTFT temperature-sensing “delta-VBE circuit” has a complementary-to-absolute-temperature (CTAT) response instead of an equivalent silicon circuit’s PTAT response. The OTFT circuit is scaled to an array format to enable surface thermal sensing applications. As Figure 2b shows, the array consists of 3x3 temperature-sensing circuit cells of 1mm2 each and is currently being characterized. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20474 Low-power wireless links are important for the development of long-term implantable neural prostheses. Furthermore, in implanted systems with many neural recording electrodes, the data rate of the wireless link will need to be quite high since each electrode typically requires at least 5kHz of bandwidth. For low-power operation, inductively-coupled near-field wireless links have shown great promise and were used to develop a power-efficient data link for biomedical implants. A bidirectional half-duplex wireless link that uses near-field inductive coupling between the implanted system and an external transceiver was designed in a 0.5-µm CMOS process. Our system minimizes power consumption in the implanted system by using impedance modulation to transmit high-bandwidth information in the uplink direction, i.e., from the implanted to the external system. We measured a data rate of 2.8Mbps at a bit error rate (BER) of <10-5 (we could not measure error rates below 10-5) and a data rate of 4.0Mbps at a BER of 10-3. Experimental results also demonstrate data transfer rates up to 300kbps in the opposite, i.e., downlink direction. Theoretical analysis of the bit error rate performance was also carried out. This analysis allowed us to theoretically predict and experimentally verify an important effect regarding the asymmetry of rising and falling edges that is inherent in impedance modulation and that contributes to bit errors. The link dissipates 2.5mW in the external system and only 100?W in the implanted system, making it among the most power-efficient inductive data links reported. The link is compatible with FCC regulations on radiated emissions. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20473 Environment-understanding technology is very vital for intelligent vehicles that are expected to respond automatically to fast- changing environments and dangerous situations. To obtain perceptual abilities, we should automatically detect static and dynamic obstacles and obtain related information such as their locations, speed, possible collision or occlusion, and other dynamic current or historic information. Conventional methods independently detect individual information, which is normally noisy and not very reliable. Instead we propose fusion-based and layered-based information-retrieval methodology as shown in Figure 1 to systematically detect obstacles and obtain their location/timing information for visible and infrared sequences. The proposed obstacle detection methodologies take advantage of the connections between different pieces of information and increase the computational accuracy of obstacle information estimation, thus improving environment understanding abilities and driving safety. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20472 Technology scaling poses challenges in designing analog circuits because of the decrease in intrinsic gain and reduced swing. An alternative to using high-gain amplifiers in the implementation of switched capacitor circuits has been proposed that replaces the amplifier with a current source and a comparator. The new comparator-based switch capacitor (CBSC) and zero-crossing-based circuit (ZCBC) techniques have been implemented in two pipelined ADC architectures at 10MHz and 200MHz and 10-bit and 8-bit accuracy, respectively. The purpose of this project is to explore the use of the ZCBC technique for very-high-precision AD converters. The goal of the project is a 100MHz, 14-bit pipelined ADC. First, we are investigating dual-phase hybrid ZCBC operation to improve the power-linearity trade off of the A/D conversion and to improve the power supply rejection. The first phase approximates the final output value, while the second phase allows the output to settle to its accurate value. Since the output is allowed to settle in the second phase, the currents through capacitors decay, permitting higher accuracy and power-supply rejection compared with standard ZCBCs. We are also developing linearization techniques for the ramp waveforms. Linear ramp waveforms require less correction in the second phase for given linearity, thus allowing faster operation. Innovative techniques for improving linearity beyond using a cascoded current source are explored, including output pre-sampling. In addition, overshoot reduction techniques will be used to improve the linearity requirements of the final phase. Alternatively, improvements in the zero-crossing detector design are pursued to lessen the linearity requirement of the ramp. Chopper stabilization will be used to reduce the effects of offsets and flicker noise. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20469 Most current ultrasound imaging systems use piezoelectric materials for the ultrasound transducer. The recent development of micro-electromechanical systems (MEMS) allowed fabrication of capacitive micromachined ultrasound transducers (CMUTs). A CMUT is a micromachined capacitor whose value changes according to the DC bias voltage or external pressure due to the physical deformation of the top plate by electrostatic force or external pressure. The major advantages of this transducer technology are the potential for integration with supporting electronic circuits, ease of fabrication, higher resolution due to small transducer size, and improved bandwidth and sensitivity. This project focuses on the front-end design of portable ultrasound systems using CMUTs. We present a conceptual block diagram of the system. Implementing an ADC at each channel input makes possible digital beam-forming in the receive (Rx) path, which enhances ultrasound image quality. To implement as many ADCs as the number of transducer channels, each ADC must consume as little power as possible, and each should be implemented in a small area. With the performance requirements at 10~20MHz sampling frequency with 8~10bits of resolution, successive approximation or zero-crossing-based (ZCB) ADC are good candidates. We are investigating these topologies to determine the optimum topology for the application. We are also exploring the control of the fire timing and pulse shape of the transmit (Tx) elements, such that beam-forming in the Tx path is possible to give a larger Rx signal. Recently, a few 2D imaging systems using CMUT as ultrasound transducers have been reported, but they do not use real-time imaging. The digital image processing block will be considered in the system level for real-time imaging. After completing the 2D ultrasound image system using a 1D transducer, we will examine the feasibility of the 3D ultrasound image system using 2D transducers. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20468 Switched-capacitor circuits can be used to implement many analog systems such as ADCs, DACs, filters, amplifiers, and integrators. In earlier phase of this research, a reconfigurable switched-capacitor system is proposed to implement different analog systems. A prototype system has been fabricated that shows basic reconfigurability to implement a pipe-lined ADC and a switched-capacitor filter. A second prototype system is being designed that utilizes highly reconfigurable blocks. We show the block diagram of such systems. The building blocks have the same functionality and can be programmed to implement a multiplier or an integrator with a reconfigurable coefficient. Such a system has many applications such as in software-defined radios and rapid prototyping of analog circuits. The design of such systems has had limited success since many different op-amp topologies are required to cover a large performance and configuration space. Recently, we proposed zero-crossing based (ZCB) circuits to design ADCs. ZCB circuits replace the op-amp in traditional switched-capacitor circuits with a combination of a current source and a zero-crossing detector. ZCB circuits are well suited for highly reconfigurable system since their power consumption scales with the operating frequency and required SNR. In addition, ZCB circuits benefit from technology scaling due to their more digital circuit-like operation in contrast to conventional op-amp based circuits. The system is designed to operate at different speeds while the power consumption is kept at the optimum level. A key challenge in design of this system is to keep the cost of programmability low in terms of additional power consumption and performance degradation. Several innovative circuit techniques have been used to address this issue. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20467 Presently, continuous monitoring of patients is very difficult and typically intrudes greatly on their lifestyles and daily routines. Methods that rely on patients measuring themselves are unreliable. An unobtrusive, minimally invasive monitoring platform would give medical scientists unprecedented access to continuous long-term data on patients. In this work, we develop a power management integrated circuit (IC) to power minimally invasive monitors using ultracapacitors. The focus of the project is to improve the evaluation and treatment of patients with a variety of movement disorders including Parkinson’s disease, restless legs syndrome, and essential tremor by allowing doctors to continuously monitor relevant biomarkers over much longer time scales and with better precision than currently possible. Because of the strict size requirements of a minimally invasive IC, a simple battery cannot be used due to the area overhead and replacement requirements. As a solution, a wirelessly rechargeable ultracapacitor network can be used as the power supply for sensing and data conversion, storage, and transmission circuitry. Ultracapacitors can carry 5-10% the energy densities of conventional battery chemistries of comparable weight. With over 105 recharge cycles possible, they are attractive candidates for implantable applications employing efficient wireless recharging schemes. A block diagram is drawn for the power management IC. Discrete ultracapacitors from Maxwell (5F, 14mmx24mmx0.5mm) are used in the initial prototype. Figure 2 demonstrates the principle of operation that achieves use of 96.8% of the initial charge. Currently, a prototype whereby the ultracapacitors are recharged using RF rectification is under development. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20466 The purpose of this project is to design an RF transmitter architecture that achieves 10 Gb/s data transfer over a 60 GHz wireless link with high power efficiency. With the availability of 7 GHz of unlicensed bandwidth centered at 60 GHz, this space has emerged as an active area of research. A number of challenges will be faced in the process of bringing this project to completion. Strong atmospheric absorption at 60 GHz lowers the signal-to-noise ratio (SNR) available at the receiver. The low SNR limits the complexity of the constellations that can be used and thus reduces the number of bits per symbol that can be encoded with the modulation strategy. Extremely fast baseband modulators will therefore be required for high data rate transmission because more symbols per second will have to be transmitted. The fundamental challenge of simultaneously obtaining good linearity and high efficiency in power amplifiers is further exacerbated at this carrier frequency, complicating transceiver design. Delivering significant power at 60 GHz requires very fast devices with high fmax and fT. This technological hurdle has been lowered with recent advances in SiGe, III-V semiconductor technology and deeply scaled CMOS. The technical approach of this project is to exploit complete co-design of the modulation strategy with a new power amplifier concept: Asymmetric Multilevel Outphasing (AMO). This architecture combines the best properties of polar transmitters and outphasing (LINC) transmitters. The power amplifier’s efficiency is improved without significantly degrading its linearity by using the combination of drain voltage modulation and rapid outphasing. A key aspect of this project will be the investigation of energy recovery as a means of further improving the transmitter’s efficiency. The use of resistance compression networks as a means of recovering the energy normally lost during outphasing will be critical. To achieve these goals, the most significant research challenges are: (1) achieving baseband modulation commensurate with 10 Gb/s transmission with the new AMO architecture and (20 designing a symbol constellation and modulation strategy that maximally exploits the architecture. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20465 The evaluation and effective treatment of patients suffering from movement disorders such as Parkinson’s disease, restless legs syndrome, and others require continuous monitoring and reliable data collection. This monitoring is a challenge due to routine movements of patients and shortcoming of the methods that rely on patients measuring themselves. To overcome these challenges the group uImplant focuses on developing an inconspicuous and minimally invasive IC system encapsulated in a bio-compatible packaging. The complete system will consist of three main areas: power management, signal collection, and data storage and communication. In this direction we have designed a customized low-power sub-threshold SRAM with on-chip features that reduce energy consumption. The SRAM is designed with the low-power application in focus. It is 32-kb, 6T bit-cell sequential read/write memory with on-chip power-saving features. Since the data gets written sequentially, the rows are powered just before they get written with valid data. After each row-write, the row written stays on for data retention. The unused rows at any given time stay unpowered, thereby saving precious energy in micro-implant applications. Further savings come by lowering the leakage currents associated with standby cells. In low-frequency sub-threshold operations, the leakage currents dominate the power consumption and are the main energy sinks. Since leakage currents are a strong function of the supply voltage, it is important to reduce the voltage to the lowest possible levels without losing the bit cell data. In this memory, the system supply voltage is stepped down using a capacitor stack that switches between a series and a parallel configuration to divide the voltage to desired voltage levels. Besides low standby leakage currents, low dynamic currents give additional savings during the write operation. The architecture enables independent floating of the supply voltage to each word in the memory. This floating trait reduces power consumption during the write operation. It also reduces the minimum size requirement on the pass transistors connecting the data input bit-lines to the bit cell during the write operation. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20464 The proposed research program has two primary goals. The first goal is to improve the evaluation and treatment of patients with diabetes and a variety of movement disorders, including Parkinson’s disease, restless legs syndrome, and essential tremor, by allowing doctors to continuously monitor relevant biomarkers over much longer time scales and with better precision than currently possible. The second goal is that the proposed implant be a platform for electronic sensory monitoring that is inexpensive and flexible and that can be used with a wide variety of sensors and for a wide variety of purposes, such as chemical sensors for monitoring blood chemistry. In this work, we develop an energy-efficient analog-to-digital converter designed to operate with a power management scheme using ultracapacitors as opposed to a battery. Two techniques are employed to save on energy for the entire system. The first is the use of an integrated capacitor that acts as a local supply for the data conversion circuit. This technique allows for us to duty-cycle the bandgap reference circuit used for power management. The second technique is to use adiabatic charging of the capacitors contained in the SAR ADC. This application is ideal for adiabatic techniques because of the low frequency of operation and the ease with which we can reclaim energy from discharging the capacitors. Building on the application, the integrated capacitor acting as a local supply allows us to reclaim energy without having to design any energy-recovery circuitry. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20463 The outphasing power amplifier dates back to the early 1930’s as an approach for the simultaneous realization of high-efficiency and high-linearity amplification. The principle of outphasing, also known as linear amplification of nonlinear components (LINC). It is based on the idea that an arbitrary input signal can be divided into two constant-amplitude, phase-modulated signals that can each be non-linearly amplified and then recombined as a vector sum to produce an output signal that is a linearly amplified version of the input. The key advantage of this approach is that each amplifier operates in an efficient albeit nonlinear mode, and yet the final output can be highly linear, breaking the usual tradeoff between efficiency and linearity. The disadvantage lies in the output-combining network: when the two amplifiers are outphased to vary the amplitude, power is wasted as heat in the isolation resistor. We describe a new outphasing energy recovery amplifier (OPERA) that replaces the isolation resistor in the conventional matched combiner with a resistance-compressed rectifier for improved efficiency. The rectifier recovers the power normally wasted in the isolation resistor back to the power supply, while a resistance compression network (RCN) reduces the impedance variation of the rectifier as the output power varies. Because the combiner requires a fixed resistance at the isolation port to ensure matching and isolation between the two outphased power amplifiers (PAs), the RCN serves to maintain high linearity as well as high efficiency in the switching-mode PAs. For demonstration, a prototype OPERA system is designed and implemented with discrete components at an operating frequency of 48MHz, delivering 20.8W peak power with 82.9% PAE. The measurement results show an efficiency improvement from 17.9% to 42.0% for a 50kHz 16-QAM signal with a peak-to-average power ratio of 6.5dB. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20461 Subsampling is recognized as an energy-efficient signal processing technique for highly digital transceivers. However, subsamplers are notorious for low SNR performance due to noise folding and for stringent requirements for anti-aliasing prefilters. This combination of faults has largely undermined their use in high-performance receivers. In transmitters, however, the situation is fundamentally different. The signal environment has fewer extreme aggressors, such as blockers, and the transmitted data is often known in advance of actual transmission. This last fact enables the use of averaging and other signal processing techniques to overcome the noise-folding problem. We show a digitally assisted subsampler, which is designed to serve as a downconversion path in adaptive predistortion transmitters with 800MHz-5.8GHz RF power amplifiers. We use digital averaging to overcome the noise-folding problems of subsampling, obtaining a final SNDR of 73.1dB for signals centered around a 2.4GHz carrier. Using quadrature subsampling, we obtain both I and Q samples from the same physical path and thereby eliminate the IQ gain mismatch. When used as part of an adaptive predistortion system, the subsampler enables an EVM improvement of 3.2% and distortion products suppression of up to 7.6dB for 802.11g signals. The subsampler IC, designed in a 90-nm CMOS process, consumes 6.0mW from a 1.2V supply. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20460 In order to increase overall power efficiency of RF power amplifiers (PAs) over a wider output power range and significantly simplify RF/analog front-end for the PAs, we designed a new outphasing transmitter architecture based on asymmetric multilevel outphasing (AMO) modulation. We compares the AMO modulation with linear amplification with nonlinear component (LINC) and multilevel (ML) LINC modulation . Independently switching the supply voltage for each PA achieves the smallest outphasing angle with AMO. Discrete supply voltage levels and low oversampling allow wideband transmission such as for WLAN, WiMAX, and 4G LTE systems. The power-efficiency improvement of AMO transmitter is compared with LINC and ML-LINC. Depending on the probability density distribution of transmission signal amplitude, each of the supply voltage levels can be optimized. For demonstration, an overall transmitter is simulated in a 65-nm CMOS process with HSUPA and WLAN signals. Compared to conventional outphasing modulation, the simulation results show an efficiency improvement from 17.7% to 40.7% for HSUPA at 25.3 dBm output power and from 11.3% to 35.5% for WLAN 802.11g at 22.8 dBm, while still meeting system linearity requirements. A compact and low-power all-digital modulator, replacing the bulky RF/analog front-end, has been designed to drive a PA in AMO transmitters. For a small silicon footprint and wideband linear operation, digital predistortion technique is applied to compensate for the mismatch existing in an open-loop direct RF phase converter. 10/19/09 http://ilp-www.mit.edu/display_project.a4d?projectId=20459 Biomedical sensors are used to measure a myriad of biopotential signals including electroencephalogram (EEG), electrocardiogram (EKG), electromyogram (EMG), and neural field potential (NFP) signals. Most of the useful information in these signals resides in the frequency range of 0.5 Hz to 1 kHz, allowing ultra-low power circuits to be used when processing them. This is critical for systems that are implanted, since energy is extremely scarce, and the lifetime of the device must be on the order of 10 years. Unfortunately, these signals are often as small as 10 µVs, and their low frequency location make them vulnerable to aggressors such as DC offset, powerline noise, and flicker noise. DC offset can result from charge accumulation at the interface between the metal electrodes and the skin, and also from amplifier offsets caused by random mismatches. While chopper stabilization has proved effective at mitigating the effects of amplifier DC offset and flicker noise, electrode DC offset cannot be removed through chopping and must be high-pass filtered at the front end of the system to prevent saturation. Powerline noise, typically at 50 or 60 Hz, is mostly a common-mode signal that requires adequate common-mode rejection. However, if there are mismatches or inductive loops in the electrodes, these aggressors can become differential-mode signals, corrupting the desired signal, and potentially saturating the system. In closed-loop deep brain stimulation systems, another aggressor arises from stimulation artifacts. In that case, the NFPs can be much smaller than stimulation artifacts placing stringent requirements on the dynamic range of the system and potentially leading to signal corruption. We propose a mixed-signal sensor interface that mitigates the effects of all of the aforementioned aggressors in an area efficient manner. Area efficiency is particularly compelling in implantable devices that use tens or hundreds of electrodes, such as neural recording systems. The proposed system uses a chopper stabilized operational amplifier with capacitive feedback to achieve accurate gain (The system is shown as single-ended for simplicity, but is implemented in a fully differential manner). We show a simplified schematic of the amplifier, including a novel input chopper that creates a switched capacitor resistance between its inputs and a reference voltage. This resistance is shown as Rp and is used to create a high-pass filter with a corner frequency well below 1 Hz, while setting the common-mode voltage of the input to a desired level. The pole frequency is actually set by the Miller-multiplied feedback capacitor Cf and is inversely proportional to the amplifier’s gain AV, allowing a reduction of many orders of magnitude in component sizes. An additional feedback path is introduced that includes the filter, ADC, DSP, and a feedback DAC. This path can be used to notch out unwanted signals such as powerline noise or stimulation artifacts before they can saturate the system. 10/19/09