I2S Masters/ Doctoral Theses

Automating the analysis of unstructured textual data, such as student course feedback, is crucial for gaining actionable insights. This project focuses on developing a sentiment analysis system leveraging the DeBERTa-v3-base model, a variant of BERT (Bidirectional Encoder Representations from Transformers), to classify feedback sentiments and generate corresponding ratings on a 1-to-5 scale.

A dataset of 100,000+ student reviews was preprocessed and fine-tuned on the model to handle class imbalances and capture contextual nuances. Training was conducted on high-performance A100 GPUs, which enhanced computational efficiency and reduced training times significantly. The trained BERT sentiment model demonstrated superior performance compared to traditional machine learning models, achieving ~82% accuracy in sentiment classification.

The model was seamlessly integrated into a functional web application, providing a streamlined approach to evaluate and visualize course reviews dynamically. Key features include a course ratings dashboard, allowing students to view aggregated ratings for each course, and a review submission functionality where new feedback is analyzed for sentiment in real-time. For the department, an admin page provides secure access to detailed analytics, such as the distribution of positive and negative reviews, visualized trends, and the access to view individual course reviews with their corresponding sentiment scores.

This project includes a comprehensive pipeline, starting from data preprocessing and model training to deploying an end-to-end application. Traditional machine learning models, such as Logistic Regression and Decision Tree, were initially tested but yielded suboptimal results. The adoption of BERT, trained on a large dataset of 100k reviews, significantly improved performance, showcasing the benefits of advanced transformer-based models for sentiment analysis tasks.

Structural health monitoring (SHM) is crucial for maintaining the safety and durability of infrastructure. To address the limitations of traditional inspection methods, this study leverages cutting-edge deep learning-based segmentation models for autonomous crack identification. Specifically, we utilized the recently launched YOLOv11 model, alongside the established DeepLabv3+ model for crack segmentation. Mask R-CNN, a widely recognized model in crack segmentation studies, is used as the baseline approach for comparison. Our approach integrates the CREC cropping strategy to optimize dataset preparation and employs post-processing techniques, such as dilation and erosion, to refine segmentation results. Experimental results demonstrate that our method—combining state-of-the-art models, innovative data preparation strategies, and targeted post-processing—achieves superior mean Intersection-over-Union (mIoU) performance compared to the baseline, showcasing its potential for precise and efficient crack detection in SHM systems.

The rapid growth of Internet of Things (IoT) technology has brought unprecedented convenience to our daily lives, enabling users to customize automation rules and develop IoT apps to meet their specific needs. However, as IoT devices interact with multiple apps across various platforms, users are exposed to complex security and privacy risks. Even interactions among seemingly harmless apps can introduce unforeseen security and privacy threats.

In this work, we introduce two innovative approaches to uncover and address these concealed threats in IoT environments. The first approach investigates hidden cross-app privacy leakage risks in IoT apps. These risks arise from cross-app chains that are formed among multiple seemingly benign IoT apps. Our analysis reveals that interactions between apps can expose sensitive information such as user identity, location, tracking data, and activity patterns. We quantify these privacy leaks by assigning probability scores to evaluate the risks based on inferences. Additionally, we provide a fine-grained categorization of privacy threats to generate detailed alerts, enabling users to better understand and address specific privacy risks. To systematically detect cross-app interference threats, we propose to apply principles of logical fallacies to formalize conflicts in rule interactions. We identify and categorize cross-app interference by examining relations between events in IoT apps. We define new risk metrics for evaluating the severity of these interferences and use optimization techniques to resolve interference threats efficiently. This approach ensures comprehensive coverage of cross-app interference, offering a systematic solution compared to the ad hoc methods used in previous research.

To enhance forensic capabilities within IoT, we integrate blockchain technology to create a secure, immutable framework for digital forensics. This framework enables the identification, tracing, storage, and analysis of forensic information to detect anomalous behavior. Furthermore, we developed a large-scale, manually verified, comprehensive dataset of real-world IoT apps. This clean and diverse benchmark dataset supports the development and validation of IoT security and privacy solutions. Each of these approaches has been evaluated using our dataset of real-world apps, collectively offering valuable insights and tools for enhancing IoT security and privacy against cross-app threats.

Sequence alignment is pivotal in both homology searches and the mapping of reads from next-generation sequencing (NGS) and third-generation sequencing (TGS) technologies. Currently, the majority of sequence alignment algorithms utilize the “seed-and-extend” paradigm, designed to filter out unrelated or nonhomologous sequences when no highly similar subregions are detected. A well-known implementation of this paradigm is BLAST, one of the most widely used multipurpose aligners. Over time, this paradigm has been optimized in various ways to suit different alignment tasks. However, while these specialized aligners often deliver high performance and efficiency, they are typically restricted to one or few alignment applications. To the best of our knowledge, no existing aligner can perform all alignment tasks while maintaining superior performance and efficiency.

In this work, we introduce a unified sequence alignment framework to address this limitation. Our alignment framework is built on the seed-and-extend paradigm but incorporates novel designs in its seeding and indexing components to maximize both flexibility and efficiency. The resulting software, the Versatile Alignment Toolkit (VAT), allows the users to switch seamlessly between nearly all major alignment tasks through command-line parameter configuration. VAT was rigorously benchmarked against leading aligners for DNA and protein homolog searches, NGS and TGS read mapping, and whole-genome alignment. The results demonstrated VAT’s top-tier performance across all benchmarks, underscoring the feasibility of using a unified algorithmic framework to handle diverse alignment tasks. VAT can simplify and standardize bioinformatic analysis workflows that involve multiple alignment tasks. 

Quantum computing has the potential to revolutionize computational problem-solving by leveraging the quantum mechanical phenomena of superposition and entanglement, which allows for processing a large amount of information simultaneously. This capability is significant in the numerical solution of complex and/or multidimensional partial differential equations (PDEs), which are fundamental to modeling various physical phenomena. There are currently many quantum techniques available for solving partial differential equations (PDEs), which are mainly based on variational quantum circuits. However, the existing quantum PDE solvers, particularly those based on variational quantum eigensolver (VQE) techniques, suffer from several limitations. These include low accuracy, high execution times, and low scalability on quantum simulators as well as on noisy intermediate-scale quantum (NISQ) devices, especially for multidimensional PDEs.

In this work, we propose an efficient and scalable algorithm for solving multidimensional PDEs. We present two variants of our algorithm: the first leverages finite-difference method (FDM), classical-to-quantum (C2Q) encoding, and numerical instantiation, while the second employs FDM, C2Q, and column-by-column decomposition (CCD). Both variants are designed to enhance accuracy and scalability while reducing execution times. We have validated and evaluated our algorithm using the multidimensional Poisson equation as a case study. Our results demonstrate higher accuracy, higher scalability, and faster execution times compared to VQE-based solvers on noise-free and noisy quantum simulators from IBM. Additionally, we validated our approach on hardware emulators and actual quantum hardware, employing noise mitigation techniques. We will also focus on extending these techniques to PDEs relevant to computational fluid dynamics and financial modeling, further bridging the gap between theoretical quantum algorithms and practical applications.

Facial emotion recognition has become a pivotal application of machine learning, enabling advancements in human-computer interaction, behavioral analysis, and mental health monitoring. Despite its potential, challenges such as data imbalance, variation in expressions, and noisy datasets often hinder accurate prediction.

This project presents a novel approach to facial emotion recognition by integrating established techniques like data augmentation and regularization with a tailored convolutional neural network (CNN) architecture. Using the FER2013 dataset, the study explores the impact of incremental architectural improvements, optimized hyperparameters, and dropout layers to enhance model performance.

The proposed model effectively addresses issues related to data imbalance and overfitting while achieving enhanced accuracy and precision in emotion classification. The study underscores the importance of feature extraction through convolutional layers and optimized fully connected networks for efficient emotion recognition. The results demonstrate improvements in generalization, setting a foundation for future real-time applications in diverse fields.

This project investigates the performance of Federated Learning Aggregation-Free (FedAF) compared to traditional federated learning methods under non-independent and identically distributed (non-IID) data conditions, characterized by Dirichlet distribution parameters (alpha = 0.02, 0.05, 0.1). Utilizing the MNIST and CIFAR-10 datasets, the study benchmarks FedAF against Federated Averaging (FedAVG) in terms of accuracy, convergence speed, communication efficiency, and robustness to label and feature skews.  

Traditional federated learning approaches like FedAVG aggregate locally trained models at a central server to form a global model. However, these methods often encounter challenges such as client drift in heterogeneous data environments, which can adversely affect model accuracy and convergence rates. FedAF introduces an innovative aggregation-free strategy wherein clients collaboratively generate a compact set of condensed synthetic data. This data, augmented by soft labels from the clients, is transmitted to the server, which then uses it to train the global model. This approach effectively reduces client drift and enhances resilience to data heterogeneity. Additionally, by compressing the representation of real data into condensed synthetic data, FedAF improves privacy by minimizing the transfer of raw data.  

The experimental results indicate that while FedAF converges faster, it struggles to stabilize under highly heterogenous environments due to limited real data representation capacity of condensed synthetic data. 

Chapter 1 introduces TinyML, emphasizing the shift from large-scale machine learning to embedded, resource-constrained devices. It explores DeepPiCar, a project demonstrating the application of TinyML principles to autonomous driving on low-cost platforms. The chapter details the motivation, setup, and fine-tuning process used to improve DeepPiCar's adaptability in changing environments using a Raspberry Pi Zero 2 W. Key findings highlight the feasibility and efficiency of on-device fine-tuning.

Chapter 2 delves into prior research that has demonstrated the effectiveness of end-to-end deep learning for robotic navigation, where the control signals are directly derived from raw sensory data. However, the majority of existing end-to-end navigation solutions are predominantly camera-based. In this chapter, we introduce TinyLidarNet, a lightweight 2D LiDAR-based end-to-end deep learning model for autonomous racing. We systematically analyze its performance on untrained tracks and computing requirements for real-time processing. We find that TinyLidarNet's 1D Convolutional Neural Network (CNN) based architecture significantly outperforms widely used Multi-Layer Perceptron (MLP) based architecture. In addition, we show that it can be processed in real-time on low-end micro-controller units (MCUs).

Chapter 3 comprehensively analyzes the variation of TinyLidarNet, an enhanced deep learning model optimized for F1TENTH autonomous racing using the ESP32S3 microcontroller and RPLiDAR A1 sensor. It is aimed at addressing challenges like size, weight, power efficiency, and environmental adaptability in the F1TENTH autonomous racing competition. Through a combination of expert driving data and the Dataset Aggregation (DAGger) technique, the model was trained to improve navigation accuracy and processing efficiency. Testing demonstrated that the vehicle could complete multiple laps on both familiar and new tracks without collisions while meeting real-time processing requirements.

In modern computer systems, sensors play a critical role in enabling a wide range of functionalities, from navigation in autonomous vehicles to environmental monitoring in smart homes. Acting as an interface between physical and digital worlds, sensors collect data to drive automated functionalities and decision-making. However, this reliance on sensor data introduces significant potential vulnerabilities, leading to various physical, sensor-enabled attacks such as spoofing, tampering, and signal injection. Sensor spoofing attacks, where adversaries manipulate sensor input or inject false data into target systems, pose serious risks to system security and privacy.

In this work, we have developed two novel sensor spoofing attack methods that significantly enhance both efficacy and practicality. The first method employs physical signals that are imperceptible to humans but detectable by sensors. Specifically, we target deep learning based facial recognition systems using infrared lasers. By leveraging advanced laser modeling, simulation-guided targeting, and real-time physical adjustments, our infrared laser-based physical adversarial attack achieves high success rates with practical real-time guarantees, surpassing the limitations of prior physical perturbation attacks. The second method embeds physical signals, which are inherently present in the system, into legitimate patterns. In particular, we integrate trigger signals into standard operational patterns of actuators on mobile devices to construct remote logic bombs, which are shown to be able to evade all existing detection mechanisms. Achieving a zero false-trigger rate with high success rates, this novel sensor bomb is highly effective and stealthy.

Our study on emerging sensor-based threats highlights the urgent need for comprehensive defenses against sensor spoofing. Along this direction, we design and investigate two defense strategies to mitigate these threats. The first strategy involves filtering out physical signals identified as potential attack vectors. The second strategy is to leverage beneficial physical signals to obfuscate malicious patterns and reinforce data integrity. For example, side channels targeting the same sensor can be used to introduce cover signals that prevent information leakage, while environment-based physical signals serve as signatures to authenticate data. Together, these strategies form a comprehensive defense framework that filters harmful sensor signals and utilizes beneficial ones, significantly enhancing the overall security of cyber systems.

In modern computer systems, sensors play a critical role in enabling a wide range of functionalities, from navigation in autonomous vehicles to environmental monitoring in smart homes. Acting as an interface between physical and digital worlds, sensors collect data to drive automated functionalities and decision-making. However, this reliance on sensor data introduces significant potential vulnerabilities, leading to various physical, sensor-enabled attacks such as spoofing, tampering, and signal injection. Sensor spoofing attacks, where adversaries manipulate sensor input or inject false data into target systems, pose serious risks to system security and privacy.

In this work, we have developed two novel sensor spoofing attack methods that significantly enhance both efficacy and practicality. The first method employs physical signals that are imperceptible to humans but detectable by sensors. Specifically, we target deep learning based facial recognition systems using infrared lasers. By leveraging advanced laser modeling, simulation-guided targeting, and real-time physical adjustments, our infrared laser-based physical adversarial attack achieves high success rates with practical real-time guarantees, surpassing the limitations of prior physical perturbation attacks. The second method embeds physical signals, which are inherently present in the system, into legitimate patterns. In particular, we integrate trigger signals into standard operational patterns of actuators on mobile devices to construct remote logic bombs, which are shown to be able to evade all existing detection mechanisms. Achieving a zero false-trigger rate with high success rates, this novel sensor bomb is highly effective and stealthy.

Our study on emerging sensor-based threats highlights the urgent need for comprehensive defenses against sensor spoofing. Along this direction, we design and investigate two defense strategies to mitigate these threats. The first strategy involves filtering out physical signals identified as potential attack vectors. The second strategy is to leverage beneficial physical signals to obfuscate malicious patterns and reinforce data integrity. For example, side channels targeting the same sensor can be used to introduce cover signals that prevent information leakage, while environment-based physical signals serve as signatures to authenticate data. Together, these strategies form a comprehensive defense framework that filters harmful sensor signals and utilizes beneficial ones, significantly enhancing the overall security of cyber systems.

Quantum computing emerges as a promising direction for high-performance computing in the post-Moore era. Leveraging quantum mechanical properties, quantum devices can theoretically provide significant speedup over classical computers in certain problem domains. Quantum algorithms are typically expressed as quantum circuits composed of quantum gates, or as unitary matrices. Execution of quantum algorithms on physical devices requires translation to machine-compatible circuits -- a process referred to as quantum compilation or synthesis. 

 Quantum synthesis is a challenging problem. Physical quantum devices support a limited number of native basis gates, requiring synthesized circuits to be composed of only these gates. Moreover, quantum devices typically have specific qubit topologies, which constrain how and where gates can be applied. Consequently, logical qubits in input circuits and unitaries may need to be mapped to and routed between physical qubits on the device.

 Current Noisy Intermediate-Scale Quantum (NISQ) devices present additional constraints, through their gate errors and high susceptibility to noise. NISQ devices are vulnerable to errors during gate application and their short decoherence times leads to qubits rapidly succumbing to accumulated noise and possibly corrupting computations. Therefore, circuits synthesized for NISQ devices need to have a low number of gates to reduce gate errors, and short execution times to avoid qubit decoherence. 

 The problem of synthesizing device-compatible quantum circuits, while optimizing for low gate count and short execution times, can be shown to be computationally intractable using analytical methods. Therefore, interest has grown towards heuristics-based compilation techniques, which are able to produce approximations of the desired algorithm to a required degree of precision. In this work, we investigate using Genetic Algorithms (GAs) -- a proven gradient-free optimization technique based on natural selection -- for circuit synthesis. In particular, we formulate the quantum synthesis problem as a multi-objective optimization (MOO) problem, with the objectives of minimizing the approximation error, number of multi-qubit gates, and circuit depth. We also employ fuzzy logic for runtime parameter adaptation of GA to enhance search efficiency and solution quality of our proposed quantum synthesis method.

There is an increasing trend in the number of wireless devices that is now already over 14 billion and is expected to grow to 40 billion devices by 2030. In addition, we are witnessing an unprecedented proliferation of applications and technologies with wireless connectivity requirements such as unmanned aerial vehicles, connected health, and radars for autonomous vehicles. The advent of new wireless technologies and devices will only worsen the current spectrum crunch that service providers and wireless operators are already experiencing. In this PhD study, we address these challenges through the following research thrusts, in which we consider two emerging applications aimed at advancing spectrum efficiency and high-frequency connectivity solutions.

First, we focus on effectively utilizing the existing spectrum resources for emerging applications such as networked UAVs operating within the Unmanned Traffic Management (UTM) system. In this thrust, we develop a coexistence framework for UAVs to share spectrum with traditional cellular networks by using machine learning (ML) techniques so that networked UAVs act as secondary users without interfering with primary users. We propose federated learning (FL) and reinforcement learning (RL) solutions to establish a collaborative spectrum sensing and dynamic spectrum allocation framework for networked UAVs. In the second part, we explore the potential of millimeter-wave (mmWave) and terahertz (THz) frequency bands for high-speed data transmission in urban settings. Specifically, we investigate THz-based midhaul links for 5G networks, where a network's central units (CUs) connect to distributed units (DUs). Through numerical analysis, we assess the feasibility of using 140 GHz links and demonstrate the merits of high-frequency bands to support high data rates in midhaul networks for future urban communications infrastructure. Overall, this research is aimed at establishing frameworks and methodologies that contribute toward the sustainable growth and evolution of wireless connectivity.