Engineering Commons^™

Holistic Dram Enhancements: From Intrinsic In-Memory Operations To Robust Security Mechanisms, Ranyang Zhou

Dissertations

Dynamic Random-Access Memory (DRAM) is both the performance bottleneck and a critical security boundary of modern computing systems. Its physical properties make it an attractive substrate for near-data computation—yet those same properties expose it to disturbance-based hardware attacks. This dissertation argues that these two dimensions are not independent: the architectural choices that make DRAM efficient also reshape its threat landscape. Addressing both requires a unified approach to memory architecture and security co-design.

The first part of this dissertation attacks the memory wall through four processing-in-DRAM (PIM) frameworks. ReD-LUT and LT-PIM unify lookup-table arithmetic with charge-sharing logic, achieving up to 37.8x …

Transcranial Ac Modulation Of Cerebellar Nuclear Activity In Awake Animals, Nuran Kavakli

Dissertations

Entrainment of cerebellar nuclear (CN) cells via cerebellar transcranial alternating current stimulation (ctACS) has been reported in animals under ketamine/xylazine anesthesia. Our main objective was to demonstrate modulation of CN activity in unanesthetized, freely moving animals using ctACS. Multi-channel carbon-fiber electrodes were implanted into the interpositus nucleus for recording multi-unit (MU) activity, and thin-film electrodes were implanted subcutaneously over the posterior cerebellum for stimulation. A frequency-domain-based metric was developed to quantify modulation from MU signals. The results demonstrated modulation in a wide range of frequencies (4 Hz-300 Hz) as in anesthetized animals. In contrast, the amplitude of the peak in …

Deep Learning Approaches For Ctenophore Identification And Tracking, Anagha Bharadwaj

Theses

Ctenophores are translucent marine organisms with nearly invisible tentacles and pose significant challenges due to their transparent morphology and ambiguous structural features. This research addresses the classification and tracking of these organisms and evaluates the performance of current computer vision models under sparse-data environments.

A dataset from the NJIT Life History Lab consisting of microscopic laboratory videos and photographs of different growth stages is used to train and assess a number of convolutional neural network designs, including VGG16, ResNet, BioCLIP2, YOLO, and DeepLabCut. Additionally, a web-based interface is developed to evaluate expert-labeled ground truth with the model's performance.

The findings …

Autonomous Exploration Of An Environment With Static Obstacles Using A Ppo Agent, Brandon Knight

Theses

Research in autonomous exploration has created many effective algorithms that have been tested and proven to work in many different virtual and physical environments. Many optimizations have also been developed to reduce computational effort and increase exploration speed.

However, despite optimizations, these algorithms can still require considerable computational effort and time to explore even small environments. To obtain further improvements in computation and exploration speed, a reinforcement learning agent using actor-critic style proximal policy optimization (PPO) is trained to explore various environments efficiently, then compared to an algorithm using contemporary exploration methods.

Testing is performed in virtual environments with ideal …

Retinomorphic Mid-Wave Infrared In-Sensor Processing Engine: Device-To-Architecture Co-Design, Hemalatha Nagaraju

Theses

Conventional frame-based CMOS image sensors acquire full-frame pixel data at discrete time intervals, resulting in substantial spatial redundancy and loss of temporal information between frames. The repeated conversion and transfer of redundant pixel data increases bandwidth and power consumption in machine vision systems. Retinomorphic sensing architectures address these limitations by enabling programmable, analog-domain processing directly at the sensor interface. A compact behavioral model of the PbSe device is developed in HSPICE based on calibrated TCAD simulation data to capture gate-controlled photocurrent modulation under varying illumination and gate bias conditions. Error analysis is performed to quantify the deviation between TCAD-generated photocurrent …

Developing A Framework For Mooc Dropout: The Role Of Utilitarian And Hedonic Values In Student Retention, Bipllab Roy, Mave D'Souza, Madhura Laghane

Northeast Journal of Complex Systems (NEJCS)

Massive Open Online Courses (MOOCs) have expanded access to higher education but continue to face persistently high dropout rates, raising concerns about their long‑term effectiveness and sustainability. This study develops and empirically tests a structural framework that links utilitarian values (perceived usefulness, certificate value, time flexibility), hedonic values (enjoyment, variety and novelty, personal interest alignment), and individual characteristics (goal orientation, self‑efficacy, motivation type) to MOOC student retention. Data were collected through a structured questionnaire administered to 200 MOOC learners from Christ University, Lavasa Campus, and analyzed using Structural Equation Modelling (SEM) in AMOS. The results show that goal orientation, self‑efficacy …

Modulating Electronic Structure With Linearly Fused Pyrazine Units For High-Voltage And Stable Zinc-Organic Batteries Cathode, Min-Jian Zhao, Li-Bin Zhang, Jin-Tao Wang, Kun Ding, Hai-Mei Liu, Yong-Gang Wang

Journal of Electrochemistry

High-voltage n-type organic cathode materials are critical for constructing zinc-organic batteries (ZOBs) with high energy density and long cycle life. However, the intrinsically unfavorable electronic structures and relatively high LUMO energy levels of most n-type materials often lead to sluggish kinetics, high solubility, and suboptimal discharge voltages (< 0.8 V). Here, we design a small molecule, quinoxalino[2’,3’:5,6]pyrazino[2,3-f][1,10]phenanthroline (DPQP), as a ZOB cathode by introducing locally electron-deficient motifs into the conjugated backbone of aromatic compounds. The linearly fused pyrazine units extending the pyrazine–benzene framework effectively optimize the electronic structure, thereby significantly enhancing the discharge voltage. Meanwhile, the expanded π-conjugated plane suppresses dissolution and accelerates charge-transfer kinetics. Benefiting from these features, the DPQP electrode exhibits an exceptional increase in average operating voltage from 0.61 V to 1.07 V (vs. Zn²⁺/Zn) at 0.1 A·g^–1, with an overpotential of only 140 mV. Notably, no discernible voltage decay occurs as the current density increases, indicating rapid and highly reversible redox kinetics. Furthermore, the DPQP cathode delivers outstanding cycling stability, maintaining over 2000 h of continuous operation at 0.1 A·g^{–1 …}

(Co,Ni,Mn,Cu,Zn)O High-Entropy Oxide Nanotubes As Efficient Bifunctional Electrocatalyst For Oxygen Evolution And Hydrazine Oxidation Reactions, Pan-Yan Chen, Wan-Wan Wu, Heng Bian, Wei-Wei Li, Xin-Sheng Zhao, Lu Wei

Journal of Electrochemistry

High-entropy oxides (HEOs) present significant scientific challenges in both design and synthesis due to their multielement and high-entropy nature, which involves complex combinations of multiple metal cations and oxygen anions, typically arranged in equimolar ratios to achieve structural stability. Herein, one-dimensional (Co,Ni,Mn,Cu,Zn)O high-entropy oxide nanotubes (HEO-NTs) are fabricated by means of a gradient electrospinning strategy with a tailored polyvinyl alcohol (PVA) molecular weight distribution and controlled pyrolysis. Benefiting from the HEO features and the synergistic effect of multicomponent sites, the as-synthesized (Co,Ni,Mn,Cu,Zn)O HEO-NTs exhibit exceptional bifunctional electrocatalytic activity for the oxygen evolution and hydrazine oxidation reactions (OER/HzOR). This study offers …

Exploring Maximal Length Cellular Automata To Generate Primitive Polynomials In Gf(2), Sumit Adak, Subhrajit Deb, Anurag Ghosh, Angshuman Roy, Souvik Roy

Northeast Journal of Complex Systems (NEJCS)

We present a simple method that uses cellular automata (CAs) to find primitive polynomials over GF(2). We used maximal length CAs as tools to generate primitive polynomials. It is usually very difficult to find maximal length CAs or primitive polynomials since they require exponential time, and there is no linear time method. However, in our work, given an n-size specific sequence of CA with reasonable probability, our technique computes a cycle of length at most 2^n-1 (maximal length) in O(n) time. The characteristic polynomials of synthesized maximal length CAs are claimed to be primitive since it was previously established that …

Decision Making At Triage Classification Using Svm With Smote Technique, Mehanas Shahul, Pushpalatha Kp

Northeast Journal of Complex Systems (NEJCS)

The efficient functioning of triage gates in overcrowded emergency departments (EDs) occurs in the context of the complex adaptive system (CAS) framework, where diverse system elements – patients, medical personnel, resources, patients’ inflow patterns, and patients themselves – simultaneously and dynamically influence the decision process. This study addresses the automated incorporation of machine learning triage algorithms as part of the system triage process to support automated classified risk-level recognition based on a limited set of vital signs. Patients are dynamically subsumed under high and low-risk categories enhanced by sensitivity, which enables optimal diagnosis and triage response to the critical clinician …

Uncovering Discrete States From Multimodal Psychophysiological Data Using Gaussian Latent Dirichlet Allocation (Glda), Congyu Wu, Aaron Fisher, David Schnyer

Northeast Journal of Complex Systems (NEJCS)

In this article we explore and validate the utility of an unsupervised probabilistic model, Gaussian Latent Dirichlet Allocation (GLDA), for discovering discrete states from repeated, multimodal psychophysiological samples collected from multiple individuals. Psychology and medical research heavily involves measuring potentially related but individually inconclusive variables from a cohort of participants to derive diagnosis, necessitating clustering analysis for state identification. Traditional probabilistic clustering models such as Gaussian Mixture Model (GMM) assume a global mixture of component distributions, which may not be realistic for observations from different patients. The GLDA model borrows the individual-specific mixture structure from a popular topic model Latent …

Les Grands Chantiers De La Transition Énergétique Au Maroc: Le Cas De Dakhla-Oued Eddahab Entre Réalisations, Facteurs De Réussite Et Perspectives, Limam Boussif, Jalila Ait Soudane

Journal of Maya Heritage

Résumé: Cet article examine la dynamique de la transition énergétique au Maroc à travers une analyse descriptive et analytique du cas de Dakhla-Oued Eddahab. Dans un contexte marqué par l’accélération des politiques de décarbonation à l’échelle mondiale, le Maroc s’impose comme un acteur stratégique en matière de développement des énergies renouvelables et de l’hydrogène vert. L’étude met en lumière les principaux chantiers structurants engagés dans la région de Dakhla, notamment le développement des énergies éolienne et solaire, les projets de production d’hydrogène vert, le dessalement de l’eau de mer, ainsi que la réalisation du port Dakhla Atlantique. L’analyse montre que …

Asymmetric Opinion Formation Of Emotional Excitable Agents, Irene Ferri, Emanuele Cozzo, Aleix Nicolás-Olivé, Albert Díaz-Guilera, Luce Prignano

Northeast Journal of Complex Systems (NEJCS)

The bounded confidence model represents a widely adopted framework for modeling opinion dynamics wherein actors have a continuous-valued opinion and interact and approach their positions in the opinion space only if their opinions are within a specified confidence threshold. Here, we propose a novel framework where the confidence bound is determined by a decreasing function of their emotional arousal, an additional independent variable distinct from the opinion value. Additionally, our framework accounts for agents' ability to broadcast messages, with interactions influencing the timing of each other's message emissions. Our findings underscore the significant role of synchronization in shaping consensus formation. …

Characteristics And Implementation Paths Of Goal For Building Energy Powerhouse, Liye Xiao, Jiaofeng Pan, Xiaojiong Wang, Deqiang Sun, Mingliang Qi, Jianlei Mo, Huimin Li, Ting Wang, Jie Yang, Jie Lin, Yuchao Wang, Haiting Chen

Bulletin of Chinese Academy of Sciences (Chinese Version)

China is a major energy-consuming country, and ensuring effective energy supply is one of the core tasks for promoting national development and national rejuvenation. To guarantee national energy supply and energy security, and to vigorously develop and utilize clean and low-carbon energy, the Outline of the 15th Five-Year Plan for National Economic and Social Development of the People’s Republic of China clearly states: “We will thoroughly implement the new energy security strategy, accelerate the construction of a clean, low-carbon, safe, and efficient new energy system, and build a strong energy nation. We will promote the safe, reliable, and orderly replacement …

Visualization And Marker-Less Tracking Of User-Defined Pre-Processed Mri Articulator Data Using Deep Learning, Michael De George

Student Theses

This thesis presents a comprehensive framework for the automated tracking and visualization of articulatory movements based on magnetic resonance imaging (MRI) data. A well-known data analysis tool for markerless pose estimation, known as DeepLabCut, is investigated for this purpose. The performance of this tool is enhanced through the design and implementation of a pre-processor. DeepLabCut is a markerless pose estimation toolbox based on deep learning, which overcomes the issue of making manual annotations frame-by-frame. Limitations from manually marking the MRI images are addressed by implementing transfer learning with convolutional neural networks to achieve accurate, user-defined articulator tracking without markers. Current …

The Impact Of Optimization Approximation Algorithms On The Performance Of The Bht-Qaoa, Ali Al-Bayaty, Marek Perkowski

Electrical and Computer Engineering Faculty Publications and Presentations

This article investigates the performance impact of five classical optimization approximation algorithms on our previously introduced quantum search algorithm, termed the Boolean–Hamiltonians Transform for Quantum Approximate Optimization Algorithm (BHT-QAOA), to effectively search for all best-approximated solutions for Boolean-based problems. These optimization approximation algorithms are BFGS, L-BFGS-B, SLSQP, COBYLA, and COBYQA. Their performance impact is evaluated and compared using two proposed performance metrics—(i) the final number of function evaluations (the lower numbers denote the best optimization approximation algorithms) and (ii) the final quality of qubit measurements (the higher values indicate all best-approximated solutions were found for a problem). Arbitrary classical Boolean …

Evaluation Of Intracavity Electromagnetic Field Probes And Sources: Implications For Shielding Effectiveness, Joseph Anthony Ferreri

Electrical and Computer Engineering ETDs

To measure the electric field in a reverberant cavity, a small, minimally invasive probe is required. Common solutions include electrically small surface mounted monopole antennas, B-dots, and D-dots. To obtain an accurate field measurement with a particular probe, it is necessary to characterize it to compensate for its ability to convert electric field into voltage which requires a gauge factor known as effective height. The characterization process is straight forward in open space on a ground plane but requires more insight when in situ in a reverberant cavity. This work adapts ground plane probe characterization methods for cavity measurements, facilitating …

Self-Supervised Spoofing Detection, David S. Choi

Electrical and Computer Engineering ETDs

Global Navigation Satellite Systems (GNSS) are vulnerable to spoofing attacks that can mislead receivers with counterfeit signals. Traditional detection techniques, such as antenna-based, encryption based, and signal processing approaches, often face limitations in adaptability, computational cost, or reliance on predefined thresholds. Supervised machine learning models, while powerful, require large labeled datasets and struggle to generalize to unseen spoofing scenarios. In this work, we propose a self-supervised spoofing detection framework based on Adaptive Sparse Gaussian Processes (ASGP). The method predicts incoming GNSS features using past observations and identifies spoofing as anomalous deviations in the prediction residuals. Unlike supervised approaches, ASGP adapts …

Millimeter-Wave Antenna Gain Enhancement Through Stacked Planar Substrates, Zachary Bergstedt

Electrical and Computer Engineering ETDs

This work presents a new wideband millimeter-wave (mmWave) and sub-terahertz antenna design with flexible directivity through the integration of stepped horn antennas and transverse substrate integrated waveguide (SIW) slots for radar and communication applications. The work gives a theoretical and analytical basis for this filter-inspired approach to improving bandwidth and directivity, and presents design and results for a standalone stepped horn, a Ka-band antenna with a solid stepped horn and SIW feed, and W-band antennas with empty SIW feeds and stepped horns manufactured out of multiple planar layers. The realized antennas show bandwidth up to 40% and gain up to …

Applications And Comparisons Of Machine Learning Methods In Ultra-Fast Laser Control, Aasma Aslam

Electrical and Computer Engineering ETDs

This dissertation demonstrates the applications and comparative analyses of machine learning methods in ultrafast laser control. By learning the relationship between the system’s input parameters and output pulse characteristics, the performance of a laser can be significantly improved. In this work, the results are presented in two stages by utilizing data from the femtosecond laser system. The first stage concerns two neural networks, named NN1 (fitrnet) and NN2 (feedforwardnet). The second stage, which extended with five different models, namely the linear regression (fitlm), the support vector machine (SVM), the Gaussian process regression (GPR), the boosted tree (fitrensemble), and LASSO (fitrlinear), …

Quantum Control Protocols For Robust Quantum Computing, Leeseok Kim

Electrical and Computer Engineering ETDs

The fundamental goal of quantum computing is to precisely control quantum systems to perform meaningful tasks, including implementing high-fidelity quantum gates for reliable quantum computation and accurately simulating complex quantum many- body dynamics. In this dissertation, we develop improved quantum control protocols for three distinct objectives, quantum error suppression, quantum optimal control, and analog quantum algorithms, achieving performance beyond standard approaches. First, we introduce new dynamical decoupling protocols, including both determin- istic and randomized constructions, that can substantially outperform conventional deterministic sequences. We then extend the randomized approach to dynamically corrected gates. Second, we propose a randomized quantum optimal control …

Fault Tolerant Quantum Computing With Lower Overhead, Benjamin E. Anker

Electrical and Computer Engineering ETDs

Quantum computation promises asymptotic speedups over classical algorithms, but realizing these advantages requires overcoming the noisiness of quantum hardware. Although fault-tolerant error correction can allow for reliable quantum computation even using unreliable components, the resource overheads required can substantially erode the asymptotic performance gains. This dissertation focuses on constructing and optimizing fault-tolerant procedures with lower overhead than previous methods were capable of. We present new frameworks for fault-tolerant syndrome extraction using flag gadgets with exponentially reduced ancilla requirements, explicit measurement schedules that achieve asymptotically fewer measurements than stabilizer generators, and a general method for making arbitrary Clifford circuits fault tolerant. …

Alternative Positioning, Navigation, And Timing In Global Navigation Satellite System Denied Environments, Joshua R. Atencio

Electrical and Computer Engineering ETDs

Global Navigation Satellite Systems (GNSS) provide the majority of critical positioning, navigation, and timing (PNT) services for civilian, commercial, and military applications. However, GNSS is vulnerable to service denial from spoofing and jamming from adversaries and environmental obstruction. These vulnerabilities highlight the need for resilient Alternative PNT (APNT) methods. This dissertation investigates APNT frameworks operating in GNSS denied environments. We develop coalition formation and matching theoretic models that allow users APNT services from anchor nodes under resource constraints and in adversarial or emergency conditions. The proposed frameworks optimize positioning accuracy, network utility, and system stability while accounting for geometric dilution …

Hierarchical Safe Reinforcement-Learning Framework For Mission-Aware, Edge-Enabled Multi-Uav Iot Networks, Abee F. Alazzwi

Electrical and Computer Engineering ETDs

This Ph.D. dissertation presents a unified Hierarchical Safe Reinforcement Learning (HSRL) framework for mission-aware, edge-enabled multi-UAV Internet of Things (IoT) networks. The work addresses the need for autonomous aerial infrastructures capable of delivering low-latency communication, scalable edge computation, and provably safe operation in dynamic environments. The dissertation develops three primary contributions. First, it formulates longhorizon drone base station placement and load balancing as a strategic actor–critic learning problem, enabling proactive adaptation to spatiotemporal demand variations. Second, it introduces a mission-aware multi-agent reinforcement learning controller for coordinated mobility, sensing, and computation offloading under latency and energy constraints. Third, it integrates a …

Calibration Methodology, Diagnostic Performance, And Redesign Of The Current Monitors On The Z Machine At Sandia National Laboratories, Stacie Hernandez

Electrical and Computer Engineering ETDs

A proper evaluation of the current diagnostics fielded in the inner-MITL region of the Z facility in 3D simulation models had not been performed until now. The evaluation of the current monitors has brought insight to their performance in a new view that has led to discoveries. The B-dot probe was the current diagnostic-of-choice since before the refurbishment of the Z facility [1] and until the development of the Inductively Driven Transmission Line (IDTL) current diagnostic [2]. Experimental data has shown that the IDTL can produce cleaner signals and it is more robust than conventional B-dots. Simulation (modeled using COMSOL …

Statistical And Spectral Theory For Spatially Correlated Random Aperiodic Antenna Arrays, Thomas Edward Christian Jr

Electrical and Computer Engineering ETDs

Aperiodic phased arrays enable beam steering, interference suppression, and spectrum efficiency for 6G, radar, biomedical imaging, and distributed sensing. Minimum inter-element spacing and keep out zones induce spatial correlation, violating the i.i.d. element-position assumption behind classical probabilistic random array theory. This dissertation develops a unified probabilistic spectral framework for correlated (non-i.i.d.) arrays. Second moment power pattern analysis incorporates the pair-correlation function  and structure factor , recovering the i.i.d. limit when  and accommodating unequal excitations. Side lobe and main lobe fields deviate from Rayleigh/Exponential and are modeled by weighted Nakagami and Gamma-mixture distributions, parameterized via Monte Carlo. The blue noise spectral …

Novel Algorithmic Methods For Random Telegraph Noise Detection And Characterization In Electronic Devices, Victor Darie Pepel

Electrical and Computer Engineering ETDs

Random telegraph noise (RTN) produces discrete stochastic fluctuations in nanoscale semiconductor devices and increasingly limits performance and reliability as dimensions scale. This dissertation introduces three algorithmic contributions enabling automated and accurate RTN characterization across diverse devices and operating conditions. First, a computationally efficient histogram-based detection algorithm enables rapid identification of RTN in large focal plane array datasets for statistically robust defect analysis. Second, a frequency decomposition framework separates slow and fast RTN components, extending the range of extractable time constants and reducing estimation error in multi-trap signals obscured by background noise. Third, to address the lack of standardized RTN metrics, …

Adapting Independent Large-Scale Pretrained Models For Human Action Recognition, Selen Pehli̇van

Turkish Journal of Electrical Engineering and Computer Sciences

Transferring knowledge from large-scale, independently pretrained image and text models to video understanding requires addressing several challenges, including maintaining generalization capabilities of models, integrating them into multimodal architectures, and fine-tuning with temporal dynamics. This study evaluates the effectiveness of parameter-efficient fine-tuning (PEFT) techniques in transferring pretrained knowledge from two independent models for video action recognition within a simple, streamlined multimodal fusion pipeline. Specifically, we adapt CLIP as the text branch and DINOv2 as the image branch, keeping both backbones frozen to preserve their pretrained robustness, while introducing lightweight, task-specific modules to adapt and fuse the branches with temporal dynamics. A …

Accurate Diagnosis Of Diseases By A Novel Ai Pipeline Based On Feature Extraction, Feature Ranking, And Feature Selection From Medical Images, Tuğba Nur Bozkurt, Mehmet Emi̇n Yüksel

Turkish Journal of Electrical Engineering and Computer Sciences

The rapid growth of the global population has led to a substantial increase in the number of patients, while the availability of healthcare professionals has not expanded at a comparable rate. This imbalance highlights the urgent need for efficient and reliable computer-aided decision support systems that can reduce clinical workload while maintaining high diagnostic accuracy. In this study, a novel and systematically integrated artificial intelligence-based pipeline is proposed for medical image classification, combining statistical significance-driven feature ranking with evolutionary feature selection in a unified framework. The proposed pipeline consists of four sequential stages: feature extraction, ranking, selection, and classification. Features …

Visual Interpretability Of Multimodal Tissue Perfusion Classification Using Grad-Cam And Saliency Maps, Metehan Zorluoglu

UNLV Theses, Dissertations, Professional Papers, and Capstones

Accurate identification of the tissue perfusion phase from hand images can aid doctors in decision-making with non-invasive techniques. The present study proposes a multimodal deep learning model for classifying the tissue perfusion phase using infrared, thermal, and visible spectrum images of the human hand. The proposed model consists of various preprocessing techniques such as manipulation, homography alignments, and masking. The significant contribution of this thesis is the interpretability analysis of deep learning models, achieved through the analysis of saliency maps and the Gradient-weighted Class Activation Mapping (Grad-CAM) methods. The purpose of this method is to find out how the convolutional …