Electrical & Computer Engineering Department Theses and Dissertations

Permanent URI for this collection

https://hdl.handle.net/1805/2087

Information about the Purdue School of Engineering and Technology Graduate Degree Programs available at IUPUI can be found at: http://www.engr.iupui.edu/academics.shtml

Browse

Now showing 1 - 10 of 242

Geocasting-based Traffic Management Message Delivery Using C-V2X
(2024-08) Mathew, Abin; Chen, Yaobin; Li, Feng; King, Brian
Cellular-Vehicle to Everything or C-V2X refers to vehicles connected to their surroundings using cellular based networks. With the rise of connected vehicles, C-V2X is emerging as one of the major standards for message transmission in automotive scenarios. The project aims to study the feasibility of C-V2X-based message transmission by building a prototype system, named RampCast, for transmitting traffic information from roadside message boards to vehicles. The RampCast framework would also implement geocasting-based algorithms to deliver messages to targeted vehicles. These algorithms focus on improving location-based message delivery using retransmission and prioritization strategies. The messages used for transmission are selected from the 511 web application built by INDOT, which contains the live traffic information for the state of Indiana which includes Travel Time information, Crash Alerts, Construction Alerts etc. The major objectives of this project consist of building the RampCast prototype, a system implementing C-V2X networks using a Software Defined Radio(SDR). The RampCast system implements a Publisher-subscriber messaging architecture with the primary actors being a Road Side Unit(RSU) and a Vehicle Onboard Unit(OBU). A data store containing traffic messages sourced from the 511 API is set up to be the input to the RampCast system. An end-to-end message transmission pipeline is built that would implement message transmission algorithms on the RSU and OBU side. Finally, the performance of message transmission on the RampCast system is evaluated using a metrics-capturing module. The system was evaluated on a test track in Columbus, Indiana. The performance metrics of the system were captured and analyzed, and the system met the key performance indicators for Latency, Packet Delivery Rate, and Packet Inter-reception Rate. The results indicate the satisfactory performance of the C-V2X standard for message transmission in the RampCast traffic guidance scenarios.
Large Language Models for Unsupervised Keyphrase Extraction and Biomedical Data Analytics
(2024-08) Ding, Haoran; Luo, Xiao; King, Brian; Zhang, Qingxue; Li, Lingxi
Natural Language Processing (NLP), a vital branch of artificial intelligence, is designed to equip computers with the ability to comprehend and manipulate human language, facilitating the extraction and utilization of textual data. NLP plays a crucial role in harnessing the vast quantities of textual data generated daily, facilitating meaningful information extraction. Among the various techniques, keyphrase extraction stands out due to its ability to distill concise information from extensive texts, making it invaluable for summarizing and navigating content efficiently. The process of keyphrase extraction usually begins by generating candidates first and then ranking them to identify the most relevant phrases. Keyphrase extraction can be categorized into supervised and unsupervised approaches. Supervised methods typically achieve higher accuracy as they are trained on labeled data, which allows them to effectively capture and utilize patterns recognized during training. However, the dependency on extensive, well-annotated datasets limits their applicability in scenarios where such data is scarce or costly to obtain. On the other hand, unsupervised methods, while free from the constraints of labeled data, face challenges in capturing deep semantic relationships within text, which can impact their effectiveness. Despite these challenges, unsupervised keyphrase extraction holds significant promise due to its scalability and lower barriers to entry, as it does not require labeled datasets. This approach is increasingly favored for its potential to aid in building extensive knowledge bases from unstructured data, which can be particularly useful in domains where acquiring labeled data is impractical. As a result, unsupervised keyphrase extraction is not only a valuable tool for information retrieval but also a pivotal technology for the ongoing expansion of knowledge-driven applications in NLP. In this dissertation, we introduce three innovative unsupervised keyphrase extraction methods: AttentionRank, AGRank, and LLMRank. Additionally, we present a method for constructing knowledge graphs from unsupervised keyphrase extraction, leveraging the self-attention mechanism. The first study discusses the AttentionRank model, which utilizes a pre-trained language model to derive underlying importance rankings of candidate phrases through self-attention. This model employs a cross-attention mechanism to assess the semantic relevance between each candidate phrase and the document, enhancing the phrase ranking process. AGRank, detailed in the second study, is a sophisticated graph-based framework that merges deep learning techniques with graph theory. It constructs a candidate phrase graph using mutual attentions from a pre-trained language model. Both global document information and local phrase details are incorporated as enhanced nodes within the graph, and a graph algorithm is applied to rank the candidate phrases. The third study, LLMRank, leverages the strengths of large language models (LLMs) and graph algorithms. It employs LLMs to generate keyphrase candidates and then integrates global information through the text's graphical structures. This process reranks the candidates, significantly improving keyphrase extraction performance. The fourth study explores how self-attention mechanisms can be used to extract keyphrases from medical literature and generate query-related phrase graphs, improving text retrieval visualization. The mutual attentions of medical entities, extracted using a pre-trained model, form the basis of the knowledge graph. This, coupled with a specialized retrieval algorithm, allows for the visualization of long-range connections between medical entities while simultaneously displaying the supporting literature. In summary, our exploration of unsupervised keyphrase extraction and biomedical data analysis introduces novel methods and insights in NLP, particularly in information extraction. These contributions are crucial for the efficient processing of large text datasets and suggest avenues for future research and applications.
Enhanced 3D Object Detection and Tracking in Autonomous Vehicles: An Efficient Multi-Modal Deep Fusion Approach
(2024-08) Kalgaonkar, Priyank B.; El-Sharkawy, Mohamed; King, Brian S.; Rizkalla, Maher E.; Abdallah, Mustafa A.
This dissertation delves into a significant challenge for Autonomous Vehicles (AVs): achieving efficient and robust perception under adverse weather and lighting conditions. Systems that rely solely on cameras face difficulties with visibility over long distances, while radar-only systems struggle to recognize features like stop signs, which are crucial for safe navigation in such scenarios. To overcome this limitation, this research introduces a novel deep camera-radar fusion approach using neural networks. This method ensures reliable AV perception regardless of weather or lighting conditions. Cameras, similar to human vision, are adept at capturing rich semantic information, whereas radars can penetrate obstacles like fog and darkness, similar to X-ray vision. The thesis presents NeXtFusion, an innovative and efficient camera-radar fusion network designed specifically for robust AV perception. Building on the efficient single-sensor NeXtDet neural network, NeXtFusion significantly enhances object detection accuracy and tracking. A notable feature of NeXtFusion is its attention module, which refines critical feature representation for object detection, minimizing information loss when processing data from both cameras and radars. Extensive experiments conducted on large-scale datasets such as Argoverse, Microsoft COCO, and nuScenes thoroughly evaluate the capabilities of NeXtDet and NeXtFusion. The results show that NeXtFusion excels in detecting small and distant objects compared to existing methods. Notably, NeXtFusion achieves a state-of-the-art mAP score of 0.473 on the nuScenes validation set, outperforming competitors like OFT by 35.1% and MonoDIS by 9.5%. NeXtFusion's excellence extends beyond mAP scores. It also performs well in other crucial metrics, including mATE (0.449) and mAOE (0.534), highlighting its overall effectiveness in 3D object detection. Visualizations of real-world scenarios from the nuScenes dataset processed by NeXtFusion provide compelling evidence of its capability to handle diverse and challenging environments.
Unraveling the Multi-omic Network and Pathway Alterations in Alzheimer's Disease
(2024-08) Xie, Linhui; Salama, Paul; Yan, Jingwen; Rizkalla, Maher; Ben Miled, Zina; Saykin, Andrew J.
Multi-omic studies ranging from genomics, transcriptomics (e.g., gene expression) to proteomics data exploration have been widely applied to interpret findings from genome wide association studies (GWAS) of Alzheimer's disease (AD). However, previous studies examine each -omics data type individually and the functional interactions between genetic variations, genes and proteins are only used after discovery to interpret the findings, but not beforehand. In this case, multi-omic findings are likely not functionally related and therefore it is challenging for result interpretation. To handle this challenge, we present new modularity constrained least absolute shrinkage and selection operator (M-LASSO), new modularity constrained logistic regression (M-Logistic), new interpretable multi-omic graph fusion neural network model (MoFNet) and new transfer learning framework integrated graph fusion neural network model (TransFuse) to integrate prior biological knowledge to model the functional interactions of multi-omic data. These approaches aim to identify functional connected sub-networks predictive of AD. In this thesis, the intrepretable model MoFNet and TransFuse incorporate prior biological connected multi-omics network, and for the first time model the dynamic information flow from deoxyribonucleic acid (DNA) to ribonucleic acid (RNA) and proteins. While applying the proposed models on multi-omic data from the religious orders study/memory and aging project (ROS/MAP) cohort, MoFNet and TransFuse outperformed all other state-of-art classifiers. Instead of targeting individual markers, the proposed methods identified multi-omic sub-networks associated with AD. MoFNet and TransFuse, produced sub-network and pathway findings that were robustly validated in another independent cohort. These identified gene/protein networks highlight potential pathways involved in AD pathogenesis and could offer systematic overview for understanding the molecular mechanisms of the disease. Investigating these identified pathways in more detail could help uncover the mechanisms causing synaptic dysfunction in AD and guide future research into potential therapeutic targets.
Development of Data-driven and AI-powered Systems Biology Methods for Understanding Human Disease
(2024-08) Dang, Pengtao; Zhang, Chi; Salama, Paul; Cao, Sha; King, Brian; Ben-Miled, Zina
Systems biology dynamic models, which are based on differential equations, offer a flexible and accurate framework to explain physiological properties emerging from complex biochem- ical or biological systems. These models enable explicit quantification and interpretation, allowing for simulation and perturbation analysis to study biological features and their inter- actions, as well as understanding system progression and convergence under various initial conditions. However, their application in human disease systems is limited due to unknown kinetics parameters under disease conditions and a reductionist paradigm that fails to cap- ture the complexity of diseases. Meanwhile, the advent of omics technologies provides high- resolution molecular measurements from single cells and spatially resolved samples, as well as comprehensive disease-specific molecular signatures from large patient cohorts. This wealth of data holds the promise for characterizing complex biological systems, necessitating ad- vanced systems biology models and computational tools that can harness multi-omics data to reliably depict biological processes. However, this endeavor faces the challenge of nonlinear relationships between omics data and the system’s dynamic properties, such as the global or local low-rank gene expression patterns across cell types and the nonlinear complexities within transcriptional regulatory networks revealed by single-cell RNA sequencing. The overall goal of this report is to develop new computational frameworks, AI-empowered methods, and related mathematical theories to explicitly represent and approximate the dy- namics of complex biological systems by using biological omics data. Our aim is to unravel the intricacies of context-specific dynamic systems using multi-Omics data. Specifically, we solved two different but related computational tasks and enabled the first-of-its-kind methods to (1) identify local low-rank matrices from large omics data, and (2) a robust optimization strategy to approximate metabolic flux. Subsequently, we delve into the realm of data-driven and AI-powered systems biology, harnessing the power of statistical learning and artificial intelligence to approximate differential equations or their representations. This research en- deavor not only contributes to the advancement of subspace modeling but also offers insights into a wide array of complex phenomena across diverse domains, with profound implications for computational biology and beyond.
Capacitorless Power Electronics Converters Using Integrated Planar Electro-Magnetics
(2024-08) Kanakri, Haitham; Cipriano Dos Santos, Euzeli, Jr.; Rizkalla, Maher; Li, Lingxi; King, Brian
The short lifespan of capacitors in power electronics converters is a significant challenge. These capacitors, often electrolytic, are vital for voltage smoothing and frequency filtering. However, their susceptibility to heat, ripple current, and aging can lead to premature faults. This can cause issues like output voltage instability and short circuits, ultimately resulting in catastrophic failure and system shutdown. Capacitors are responsible for 30% of power electronics failures. To tackle this challenge, scientists, researchers, and engineers are exploring various approaches detailed in technical literature. These include exploring alternative capacitor technologies, implementing active and passive cooling solutions, and developing advanced monitoring techniques to predict and prevent failures. However, these solutions often come with drawbacks such as increased complexity, reduced efficiency, or higher upfront costs. Additionally, research in material science is ongoing to develop corrosion-resistant capacitors, but such devices are not readily available. This dissertation presents a capacitorless solution for dc-dc and dc-ac converters. The proposed solution involves harnessing parasitic elements and integrating them as intrinsic components in power converter technology. This approach holds the promise of enhancing power electronics reliability ratings, thereby facilitating breakthroughs in electric vehicles, compact power processing units, and renewable energy systems. The central scientific premise of this proposal is that the capacitance requirement in a power converter can be met by deliberately augmenting parasitic components. Our research hypothesis that incorporating high dielectric material-based thin-films, fabricated using nanotechnology, into planar magnetics will enable the development of a family of capacitorless electronic converters that do not rely on discrete capacitors. This innovative approach represents a departure from the traditional power converter schemes employed in industry. The first family of converters introduces a novel capacitorless solid-state power filter (SSPF) for single-phase dc-ac converters. The proposed configuration, comprising a planar transformer and an H-bridge converter operating at high frequency, generates sinusoidal ac voltage without relying on capacitors. Another innovative dc-ac inverter design is the twelve step six-level inverter, which does not incorporate capacitors in its structure. The second family of capacitorless topologies consists of non-isolated dc-dc converters, namely the buck converter and the buck-boost converter. These converters utilize alternative materials with high dielectric constants, such as calcium copper titanate (CCTO), to intentionally enhance specific parasitic components, notably inter capacitance. This innovative approach reduces reliance on external discrete capacitors and facilitates the development of highly reliable converters. The study also includes detailed discussions on the necessary design specifications for these parasitic capacitors. Furthermore, comprehensive finite element analysis solutions and detailed circuit models are provided. A design example is presented to demonstrate the practical application of the proposed concept in electric vehicle (EV) low voltage side dc-dc power converters used to supply EVs low voltage loads.
Analysis of Latent Space Representations for Object Detection
(2024-08) Dale, Ashley Susan; Christopher, Lauren; King, Brian; Salama, Paul; Rizkalla, Maher
Deep Neural Networks (DNNs) successfully perform object detection tasks, and the Con- volutional Neural Network (CNN) backbone is a commonly used feature extractor before secondary tasks such as detection, classification, or segmentation. In a DNN model, the relationship between the features learned by the model from the training data and the features leveraged by the model during test and deployment has motivated the area of feature interpretability studies. The work presented here applies equally to white-box and black-box models and to any DNN architecture. The metrics developed do not require any information beyond the feature vector generated by the feature extraction backbone. These methods are therefore the first methods capable of estimating black-box model robustness in terms of latent space complexity and the first methods capable of examining feature representations in the latent space of black box models. This work contributes the following four novel methodologies and results. First, a method for quantifying the invariance and/or equivariance of a model using the training data shows that the representation of a feature in the model impacts model performance. Second, a method for quantifying an observed domain gap in a dataset using the latent feature vectors of an object detection model is paired with pixel-level augmentation techniques to close the gap between real and synthetic data. This results in an improvement in the model’s F1 score on a test set of outliers from 0.5 to 0.9. Third, a method for visualizing and quantifying similarities of the latent manifolds of two black-box models is used to correlate similar feature representation with increase success in the transferability of gradient-based attacks. Finally, a method for examining the global complexity of decision boundaries in black-box models is presented, where more complex decision boundaries are shown to correlate with increased model robustness to gradient-based and random attacks.
Explainable AI Methods For Enhancing AI-Based Network Intrusion Detection Systems
(2024-08) Arreche, Osvaldo Guilherme; King, Brian S.; Abdallah, Mustafa; El-Sharkawy, Mohamed A.
In network security, the exponential growth of intrusions stimulates research toward developing advanced artificial intelligence (AI) techniques for intrusion detection systems (IDS). However, the reliance on AI for IDS presents challenges, including the performance variability of different AI models and the lack of explainability of their decisions, hindering the comprehension of outputs by human security analysts. Hence, this thesis proposes end-to-end explainable AI (XAI) frameworks tailored to enhance the understandability and performance of AI models in this context.The first chapter benchmarks seven black-box AI models across one real-world and two benchmark network intrusion datasets, laying the foundation for subsequent analyses. Subsequent chapters delve into feature selection methods, recognizing their crucial role in enhancing IDS performance by extracting the most significant features for identifying anomalies in network security. Leveraging XAI techniques, novel feature selection methods are proposed, showcasing superior performance compared to traditional approaches.Also, this thesis introduces an in-depth evaluation framework for black-box XAI-IDS, encompassing global and local scopes. Six evaluation metrics are analyzed, including descrip tive accuracy, sparsity, stability, efficiency, robustness, and completeness, providing insights into the limitations and strengths of current XAI methods.Finally, the thesis addresses the potential of ensemble learning techniques in improving AI-based network intrusion detection by proposing a two-level ensemble learning framework comprising base learners and ensemble methods trained on input datasets to generate evalua tion metrics and new datasets for subsequent analysis. Feature selection is integrated into both levels, leveraging XAI-based and Information Gain-based techniques.Holistically, this thesis offers a comprehensive approach to enhancing network intrusion detection through the synergy of AI, XAI, and ensemble learning techniques by providing open-source codes and insights into model performances. Therefore, it contributes to the security advancement of interpretable AI models for network security, empowering security analysts to make informed decisions in safeguarding networked systems.
Design of Ultra-Low Power FinFET Charge Pumps for Energy Harvesting Systems
(2024-08) Atluri, Mohan Krishna; Rizkalla, Maher E.; King, Brian S.; Christopher, Lauren A.
This work introduces an ultra-low-voltage charge pump for energy harvesters in biosensors. The unique aspect of the proposed charge pump is its two-level design, where the first stage elevates the voltage to a specific level, and the output voltage of this stage becomes the input voltage of the second stage. Using two levels reduces the number of stages in a charge pump and improves efficiency to get a higher voltage gain. In our measurements, this charge pump design could convert a low 85mV input voltage to a substantial 608.2mV output voltage, approximately 7.15 times the input voltage, while maintaining a load resistance of 7MΩ and a 29.5% conversion efficiency.
Lidar Based 3D Object Detection Using Yolov8
(2024-08) Menon, Swetha Suresh; El-Sharkawy, Mohamed; King, Brian; Rizkalla, Maher
Autonomous vehicles have gained substantial traction as the future of transportation, necessitating continuous research and innovation. While 2D object detection and instance segmentation methods have made significant strides, 3D object detection offers unparalleled precision. Deep neural network-based 3D object detection, coupled with sensor fusion, has become indispensable for self-driving vehicles, enabling a comprehensive grasp of the spatial geometry of physical objects. In our study of a Lidar-based 3D object detection network using point clouds, we propose a novel architectural model based on You Only Look Once (YOLO) framework. This innovative model combines the efficiency and accuracy of the YOLOv8 network, a swift 2D standard object detector, and a state-of-the-art model, with the real-time 3D object detection capability of the Complex YOLO model. By integrating the YOLOv8 model as the backbone network and employing the Euler Region Proposal (ERP) method, our approach achieves rapid inference speeds, surpassing other object detection models while upholding high accuracy standards. Our experiments, conducted on the KITTI dataset, demonstrate the superior efficiency of our new architectural model. It outperforms its predecessors, showcasing its prowess in advancing the field of 3D object detection in autonomous vehicles.

Browse

Recent Submissions