BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Department of Electrical & Computer Engineering - ECPv6.15.20//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://ece.northeastern.edu
X-WR-CALDESC:Events for Department of Electrical & Computer Engineering
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20200308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20201101T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20210314T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20211107T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20220313T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20221106T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20210903T110000
DTEND;TZID=America/New_York:20210903T120000
DTSTAMP:20260517T170240
CREATED:20210825T175854Z
LAST-MODIFIED:20210825T175854Z
UID:5137-1630666800-1630670400@ece.northeastern.edu
SUMMARY:ECE PhD Proposal Review: Nikita Mirchandani
DESCRIPTION:PhD Proposal Review: Ultra-Low Power and Robust Analog Computing Circuits and System Design Framework for Machine Learning Applications \nNikita Mirchandani \nLocation: Zoom Link \nAbstract: As the scaling of CMOS transistors has almost halted\, performance gains of digital systems have also started to stagnate. There is a renewed interest in alternate computing techniques such as in-memory computing\, hybrid computing\, approximate computing\, and analog computing. In particular\, analog computing has reemerged as a promising alternative to save power and improve performance specifically for machine-learning (ML) applications. Analog computing has better area and power efficiency when compared to their digital counterpart. Power and chip area efficiency make analog computing highly appealing for implementing deep learning algorithms on-chip\, computing circuits for the internet-of-things (IoT) devices\, and implantable and wearable biomedical devices. However\, compared to digital computing\, analog computing methods have not nearly been utilized to their fullest potential due to longstanding challenges related to reliability\, programmability\, power consumption\, and high susceptibility to variations.\nThe subject of this dissertation research is to develop robust ultra-low power analog hardware suitable for machine learning applications. First\, a robust analog design methodology is presented to address issues of variability in analog circuits. A constant transconductance design technique using switched capacitor circuits is presented. The design approach is then applied to build circuits for ML applications. An analog vector matrix multiplier (VMM) is designed to be used in the convolutional layer in an ML analog computing vision hardware platform. Computing circuits are tested as part of an image classification DNN algorithm on the MNIST dataset and can achieve a classification accuracy of 96.1%. \nThe design approach is also used to design an analog computing system architecture for detection of seizures using EEG signals. A conventional EEG monitoring system includes an analog front-end (AFE)\, ADC\, digital filtering stage\, EEG feature extraction engine\, and SVM classification. Such systems suffer from high power and chip area requirements. The corresponding analog architecture is composed of AFE amplifiers to provide gain for the incoming signal. The AFE is followed by an analog filtering stage\, where spectral power from each of the bands is used as a feature for seizure classification. The output of each filter is applied to a corresponding feature extraction circuit to continuously monitor the onset of a seizure in an ultra-lower power mode with sub-threshold analog processing. The system level architecture is first modeled to obtain classification accuracy of seizures. Simulation times for the design of such complex analog systems can be prohibitively long\, particularly when the impacts of nonidealities such as noise\, nonlinearity\, and device mismatches have to be considered at the system level. The simulation time is reduced by building accurate models of the analog blocks for faster simulations. The analog models help to define the required specifications for each block in order to achieve a specified system-level classification accuracy.\nInfrastructure circuits like oscillators and voltage regulators for the proposed SoC are presented. A 254 nW 21 kHz on-chip RC oscillator with 21.5 ppm/oC temperature stability is presented to provide stable clock source for the proposed SoC.
URL:https://ece.northeastern.edu/event/ece-phd-proposal-review-nikita-mirchandani/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20210910T140000
DTEND;TZID=America/New_York:20210910T150000
DTSTAMP:20260517T170240
CREATED:20210908T194338Z
LAST-MODIFIED:20210908T194338Z
UID:5156-1631282400-1631286000@ece.northeastern.edu
SUMMARY:ECE PhD Dissertation Defense: Wenqian Liu
DESCRIPTION:PhD Dissertation Defense: Explainable Efficient Models for Computer Vision Applications \nWenqian Liu \nLocation: Zoom Link \nAbstract: State of the art deep learning based models\, such as Convolutional Neural Networks (CNNs) and generative models\, achieve impressive results\, but with their great performance comes great complexity and opacity\, huge parametric spaces and little explainability. The criticality of model explainability and output interpretability\, manifests clearly in real-time critical decision making processes and human-centred applications\, such as in healthcare\, security and insurance. Explainability and interpretability are tackled in this thesis\, as intrinsic qualities in the model architecture as well as post-hoc improvement on existing models. In the area of frame prediction in video sequences\, we introduce DYAN\, a novel network with very few parameters\, that is easy to train and produces accurate high quality predictions. Another key aspect of DYAN is interpretability\, as its encoder-decoder architecture is designed following concepts from systems identification theory and exploits the dynamics-based invariants of the data. We also introduce KW-DYAN\, an extension of DYAN that tackles the issue of time lagging in video predictions\, by implementing a novel way of quantifying prediction timeliness and proposing a new recurrent network for adaptive temporal sequence prediction. The experimental results show the reduced lagging across datasets\, while also performing well in other metrics. In this thesis we also propose the first technique to visually explain VAEs by means of gradient-based attentions\, with methods to generate visual attentions from the learned latent space\, and also demonstrate such attention explanations serve more than just explaining VAEs. We show how these attention maps can be used to localize anomalies in images\, conducting state-of-the-art performance on multiple datasets. We also apply our technique for skin image anomaly detection and diagnosis and achieve competitive quantitative and qualitative results.
URL:https://ece.northeastern.edu/event/ece-phd-dissertation-defense-wenqian-liu/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20210914T140000
DTEND;TZID=America/New_York:20210914T150000
DTSTAMP:20260517T170240
CREATED:20210908T184332Z
LAST-MODIFIED:20210908T184332Z
UID:5151-1631628000-1631631600@ece.northeastern.edu
SUMMARY:ECE PhD Dissertation Defense: Vageeswar Rajaram
DESCRIPTION:PhD Dissertation Defense: Near-Zero Power Microelectromechanical Sensors for Large-Scale IoT Sensor Networks \nVageeswar Rajaram \nLocation: ISEC 432 \nAbstract: The Internet-of-things revolution has ushered the development of sensing technologies aimed towards establishing large-scale remote sensor networks to monitor the environment continuously and with high spatial resolution. However\, with existing sensor technologies this goal has so far been limited in terms of scalability (i.e.\, the number of sensors in a network\, areal coverage and spatial granularity). A major impeding factor is sensor power consumption: state-of-the-art remote sensor technologies need to be actively powered (i.e.\, by a battery) to continuously monitor the environment for an object of interest\, even at standby (when it is not present). This is because all signals collected by the sensor from the environment need to be processed by active signal conditioning circuits to distinguish a signal of interest from other signals. Therefore\, in applications where an event or signal of interest occurs only occasionally\, most of the battery is drained by processing irrelevant signals. The result is that as the sensor network scales up\, so do the costs and labor associated with the sensors’ battery replacements. This makes it unfeasible to deploy and maintain large numbers of sensors for any application and greatly limits the scale of sensor networks. Extremely low power consumption therefore is critical in enabling large sensor networks by reducing or even eliminating costs associated with frequent battery replacements.\nThis work describes the development of a revolutionary new sensing platform aimed at creating sensors with battery lifetimes limited only by the self-discharge of the battery itself (>10 years). The ultimate goal for the technology is to enable maintenance-free sensor nodes for truly large-scale “deploy-and-forget” sensor networks. In particular\, this work details the development of novel infrared sensors based on micro-electro-mechanical photoswitches that are capable of detecting and distinguishing specific infrared signatures associated with objects of interest (hot gases\, fire\, human body\, etc.) while remaining dormant with near-zero power consumption at standby. This unique sensor technology aims to break the paradigm of requiring a power supply to perform sensing by instead relying on the energy contained in the infrared signals emitted by the object of interest itself to perform its detection. This dissertation presents a comprehensive summary of the sensor’s design\, its capabilities\, and the various technical developments that have led this technology to evolve from a concept to a prototype near-zero power wireless infrared sensor with orders of magnitude lesser power consumption compared to the state-of-the-art.
URL:https://ece.northeastern.edu/event/ece-phd-dissertation-defense-vageeswar-rajaram/
LOCATION:432 ISEC\, 360 Huntington Ave\, Boston\, MA\, 02115\, United States
GEO:42.3396156;-71.0886534
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=432 ISEC 360 Huntington Ave Boston MA 02115 United States;X-APPLE-RADIUS=500;X-TITLE=360 Huntington Ave:geo:-71.0886534,42.3396156
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20210916T160000
DTEND;TZID=America/New_York:20210916T170000
DTSTAMP:20260517T170240
CREATED:20210908T194439Z
LAST-MODIFIED:20210908T194439Z
UID:5158-1631808000-1631811600@ece.northeastern.edu
SUMMARY:ECE PhD Proposal Review: Nirjhar Bhattacharjee
DESCRIPTION:PhD Proposal Review: Sputtered Topological Insulator/Ferromagnet Heterostructures for Energy Efficient Spintronic Device Applications \nNirjhar Bhattacharjee \nLocation: Zoom Link \nAbstract: Topological insulators (TI) are Van der Waals (VdW) layered materials which possess enormous spin-orbit coupling (SOC) strength and spin-momentum locked robust surface states. TIs in presence of time-reversal symmetry breaking magnetic order can also exhibit chiral quantum anomalous (QAH) or Axion insulator edge channels. These and myriad of other material properties predicted\, and achievable utilizing TIs\, magnetic-TIs (MTI) and TI based heterostructures can open the path towards realization of a diverse class of energy efficient spintronic devices for information processing and storage. Crystalline oriented TIs which possess topologically nontrivial properties are grown using molecular beam epitaxy (MBE) which is incompatible with industrial CMOS processes. Magnetron sputtering\, on the other hand is the CMOS industry stand thin film growth technique because of the advantages of high throughput\, large area\, and high quality thin film growth capability. In this work\, first the growth of high quality TI: Bi2Te3 thin films using CMOS compatible magnetron sputtering process is introduced. Next\, room temperature characterization of magnetic and SOT properties of TI/ferromagnet (FM) heterostructures will be presented. Finally\, fascinating magnetic properties of material systems with FM species intercalated in TIs will be reported which can possibly house exotic quantum material phases. \nBy varying process temperature between 20-250ºC\, growth of Bi2Te3 with stoichiometric composition and varying crystalline order from disordered to highly c-axis oriented VdW layered films were obtained. Using X-ray diffraction (XRD) and high-resolution transmission electron microscopy (HRTEM) imaging\, the crystalline property of the TI film was confirmed. Further\, coupling the sputtered TI films with ferromagnetic (FM) thin films surprisingly showed a giant enhancement in Gilbert damping with c-axis oriented TI which is crucial for energy efficient SOT-MRAM devices. This suggested enhancement in spin-orbit coupling strength for c-axis oriented TI thin films compared to disordered ones. Formation of interface layers because of elemental diffusion has been reported in literature. But\, literature reports on SOT characteristics have largely assumed atomically sharp interfaces between TI and FM layers. We observed crystalline order dependent interface thickness and composition in Bi2Te3/Ni80Fe20 heterostructures because of diffusion of Ni across the interface. An enhancement in damping-line SOT in crystalline ordered Bi2Te3 was observed. The spin-charge conversion efficiency was however found to be larger for granular and lowest for polycrystalline disordered Bi2Te3 samples considering the interface layers. Further\, with the intercalation of Ni in Bi2Te3\, emergence of an antiferromagnetic VdW phase was observed in Ni-intercalated Bi2Te3 interface. This AFM VdW interface resulted in a large spontaneous exchange bias in Bi2Te3/Ni80Fe20 and Bi2Te3/NiZn-Ferrite heterostructures at temperatures below ~63 K which is higher than the transition temperatures of MTIs reported in literature. Structural and chemical characterization of the Ni-intercalated Bi2Te3 showed evidence of formation of Ni-Te bonds and indicated towards formation of MTI compounds. These results open new avenues for experimental exploration of fascinating high-temperature QAH and other topologically nontrivial material phases in interfaces of industrial CMOS process compatible sputter-grown TI/FM heterostructures. Understanding the properties of these TI based material systems can lead to realization of robust energy efficient spintronic devices.
URL:https://ece.northeastern.edu/event/ece-phd-proposal-review-nirjhar-bhattacharjee/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20210920T130000
DTEND;TZID=America/New_York:20210920T140000
DTSTAMP:20260517T170240
CREATED:20210920T174407Z
LAST-MODIFIED:20210920T174931Z
UID:5172-1632142800-1632146400@ece.northeastern.edu
SUMMARY:Distinguished Lecture: Dr. Elliot Eichen
DESCRIPTION:The Institute for the Wireless Internet of Things is pleased to host a distinguished lecture on Real-time Geo-spatial Spectrum Sharing by Dr. Elliot Eichen. \nWhen: Monday\, September 20th\, 1pm. \nLocation: Zoom Link \nAbstract: New technology and new applications for wireless communications have created competition for frequency bands traditionally allocated to remote sensing and defense applications. Competition for spectrum is particularly intense in mm (and sub-mm) wave bands where the requirements for 5G/6G transmissions overlap with measurements made by passive radiometers (Earth Exploration Satellite Services – EESS) that are used for weather forecasting and as baseline data for climate models. Real-time Geospatial Spectrum Sharing (RGSS) enables EESS radiometers and 5G/6G networks to gracefully share spectrum by modifying network traffic during the time window (~ 10s of msec) that a base station (gNB) and its connected endpoints (UEs) are within the effective field of view (eFOV) of a radiometer. RGSS is based on existing network infrastructure rather than Monte-Carlo network simulations (the ITU model); it can provide better isolation between 5G/6G transmissions and EESS radiometers than the ITU’s hardware-based Out-of-Band (OOB) emission limits (e.g.\, -32 dBW/200MHz-gNB and -29 dBW/200MHz-UE) in dense urban environments\, while simultaneously enabling carriers to create larger cell sizes and use network repeaters in suburban and rural settings. In addition\, RGSS can adapt to changes in network or remote sensing technology by modifying the underlying network or EESS ecosystem descriptions (schemas). \nIn this talk\, we show that RGSS: \n\nCan prevent 5G/6G transmissions from corrupting EESS measurement data\nHas sufficient geolocation accuracy to provide a realistic solution\, based on experimental confirmation of predicted measurement pixels vs. actual measurement pixels\nApplies to all mm-wave and submm-wave bands (e.g.\, a single system can be used for all bands\, such as 24\, 51\, and 90 GHz\, although the modification time windows for each band may be different)\nEnables carriers to optimize network performance by geography and time of day\, rather than designing for the worst-case scenario across the entire network (i.e.\, avoids the ” one size fits all” OOB emissions model)\nIncludes the effect of massive Multiple-Input Multiple-Output (MIMO) beamforming antennas\nIs commensurate with existing 5G architectures and deployment models\, and\nProvides a simple mechanism to test and police compliance compared with over the air TRP OOB measurements.\n\nBio: Elliot Eichen retired as Director of R&D at Verizon in 2017\, after a 35-year career (except for 2½ years on staff at MIT) at GTE Labs\, GTE/BBN\, and Verizon Labs. From 2018-2019\, he was an IEEE-USA/AAAS congressional fellow\, which is where he became interested in spectrum management and the overlap between 5G/6G and EESS passive sensors. Dr. Eichen received a Ph.D. in Optics from The University of Arizona\, and a B.S in Physics from SUNY Stony Brook. His contributions to the technical community include associate editor of IEEE Photonics Technology Letters\, committee chair of Optical Fiber Communications (OFC)\, chair of the IEEE/OSA Optical Amplifier Conference\, Visiting Industry Professor at Tufts University\, and adjunct faculty at NEU. He has more than 40 peer-reviewed publications and about 60 US patents.
URL:https://ece.northeastern.edu/event/distinguished-lecture-dr-elliot-eichen/
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20210927T110000
DTEND;TZID=America/New_York:20210927T120000
DTSTAMP:20260517T170240
CREATED:20210920T183529Z
LAST-MODIFIED:20210920T183529Z
UID:5177-1632740400-1632744000@ece.northeastern.edu
SUMMARY:ECE PhD Proposal Review: Tianyu Dai
DESCRIPTION:PhD Proposal Review: Data-Driven Control and Estimation \nTianyu Dai \nLocation: Zoom Link \nAbstract: During the last two decades\, data-driven control (DDC) has attracted growing attention in the control community. Unlike model-based control (MBC) that first uses the collected data to identify the system\, then designs the controller according to the certainty equivalence principle\, DDC skips the system identification (SYSID) step and leads to a control law directly from data. One important feature of DDC is that some fundamental limitations of MBC such as uncertainty versus robustness\, inevitable modeling error\, and possible expensive cost of SYSID are avoided in the DDC framework. These benefits enable the researcher to design controllers with better performance and accuracy. \nThe aim of this proposal is to summarize our contributions to the DDC field. We mainly discuss the following problem: given a single trajectory of noisy data and a few priors of the model structure\, how to design a state feedback controller to stabilize the system with unknown dynamics and in addition\, to meet some performance criteria. The main idea hinges on duality theory to establish the connection between two sets\, one compatible with the noisy data\, and the second satisfying some design properties such as stability or optimality. Our main results show that for all possible systems compatible with the data\, the data-driven control law can be obtained by solving a convex optimization problem. \nThis proposal is organized as follows: to start with\, we propose a DDC framework for switched linear systems relying on the Farkas’ lemma to search for a common polyhedral control Lyapunov function using the theory of moments. Then to reduce the computational burden\, we provide another method called data-driven quadratic stabilization control for linear systems that is based on quadratic Lyapunov function. To deal with nonlinear system\, we first design data-driven controllers for polynomial systems using the dual Lyapunov theorem. Then to handle general nonlinearities\, we propose a method based on state-dependent representations. Finally\, a data-driven estimator is proposed that gives the worst-case optimal estimation of the trajectory of a switched linear system.
URL:https://ece.northeastern.edu/event/ece-phd-proposal-review-tianyu-dai/
END:VEVENT
END:VCALENDAR