University of Illinois Urbana-Champaign
Illinois Digital Environment for Access to Learning and Scholarship RepositoryNot a member yet
123813 research outputs found
Sort by
Supporting write-through caching with high-performance TSO for resilient CXL systems
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Burak Ocalan, accepted the attached license on 2025-07-10 at 09:47.The student, Burak Ocalan, submitted this Thesis for approval on 2025-07-10 at 11:59.This Thesis was approved for publication on 2025-07-11 at 14:06.DSpace SAF Submission Ingestion Package generated from Vireo submission #22481 on 2025-10-20 at 20:15:08Current multiprocessors that support the total store order (TSO) memory consistency model invariably use write-back (WB) caches. When their hardware needs to issue write-through (WT) stores as in uncached operations, they deliver dismal performance: writes to memory have to be fully serialized, often forcing the program to observe the full latency of a round-trip to memory. To solve this problem, this thesis presents a novel architecture that supports high-performance TSO for WT caching environments. The architecture, called PhasedStore, involves extending the store queue of the core with additional states. Individual WT stores operate in two phases, which allow them to fully overlap with other stores and still satisfy TSO. PhasedStore is useful in environments that require a WT cache coherence protocol. This is the case in a setting that this thesis explores: a cache-coherent CXL distributed shared memory system in the presence of compute node failures. In this environment, to simplify recovery, we support a WT directory-based protocol which guarantees that stores are propagated to the CXL memory before updating the caches. Our evaluation shows that PhasedStore is very effective. Compared to a conventional approach to enforce TSO in a WT protocol, PhasedStore reduces the average execution time of a set of parallel applications by 2.21x. Further, the applications' execution time is only 1.40x higher than with a TSO WB protocol that includes no resilience consideration
Developing a virtual reality information gap activity for collaborative language learning
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, Peter Grzywacz, accepted the attached license on 2025-07-16 at 15:01.The student, Peter Grzywacz, submitted this Thesis for approval on 2025-07-16 at 15:07.This Thesis was approved for publication on 2025-07-22 at 16:35.DSpace SAF Submission Ingestion Package generated from Vireo submission #22608 on 2025-10-20 at 20:15:21One of Virtual Reality’s affordances is the ability to hide and show synchronized information to the users. This thesis documents the research, design, development and testing of the transfer of a popular communicative ESL teaching activity, the “information gap” activity, from its physical paper and pen origins into the medium of virtual reality. This activity uses affordances, signifiers, instructions, set theory, and a game engine to create an interactive and immersive activity for language learning. The implementation of the technical components and development process are documented so that other developer-instructors can gain a viewpoint on developing ESL activities in the new medium of virtual reality with the goal of providing learners with interactive and collaborative learning experiences
Decades of resistance: method and motivation in the struggle against Narita International Airport (1964-present)
Get PDFSubmission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-20 without embargo termsThe student, David Wallace, accepted the attached license on 2025-07-22 at 03:07.The student, David Wallace, submitted this Thesis for approval on 2025-07-22 at 03:24.This Thesis was approved for publication on 2025-07-24 at 16:25.DSpace SAF Submission Ingestion Package generated from Vireo submission #22672 on 2025-10-20 at 20:15:32The Sanrizuka Struggle (1964 to Present) saw farming families unite with leftist student groups to protest against the construction of New Tokyo International Airport (now Narita Airport). I trace the project’s development, from its benign origins as a means of alleviating congestion at Tokyo’s Haneda Airport to its escalation into militant protest and violence between protesters, police and government workers. I analyze the myriad factors that contributed to the outbreak of unprecedented levels of violence during the struggle, and work to fit the protests into a larger chronology of environmental and social movements in Japan and worldwide. I conclude that the Sanrizuka Struggle represents something groundbreaking in the realm of Japanese counterculture: a concerted effort by disparate factions to resist the Liberal Democratic Party (LDP) government’s policy of “progress at any cost.” Although earlier scholars have oversimplified the protests as NIMBYist violence against an inconvenient land development project, I show how the movement was multifaceted and complex, inspired not just by local concerns, but national issues. With time, it grew to include supporters from across Japan and around the world. Hantai Dōmei protestors sought to raise awareness of injustices they saw as a threat to their way of life. The construction of New Tokyo International Airport brought the Hantai Dōmei’s farmers and leftist students together out of a sense of economic, environmental, and political necessity
Thermal performance analysis of next generation high performance computing data centers using waterfall diagrams
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Donggun Lee, accepted the attached license on 2025-06-26 at 12:41.The student, Donggun Lee, submitted this Thesis for approval on 2025-06-26 at 12:49.This Thesis was approved for publication on 2025-07-07 at 12:41.DSpace SAF Submission Ingestion Package generated from Vireo submission #22371 on 2025-10-21 at 10:05:35Thermal management in high-density computing data centers is crucial for maintaining performance, reliability, and energy efficiency. As computing power demands surge, traditional air-cooling methods face limitations in handling increasing heat loads, while liquid cooling offers a feasible solution. However, a system-level metric and analysis assessing the energy efficiency of liquid cooling remains absent. This paper introduces a novel waterfall diagram (WFD) framework to provide a comprehensive system-level analysis of liquid cooling performance. The WFD evaluates energy efficiency using total-power usage effectiveness (TUE) derived from analytical calculations. A comparative analysis of coolant distribution unit (CDU) designs identified the 3U CDU architecture as the most suitable configuration for WFD analysis due to its optimal power consumption, TUE, and power density. The WFD analysis also incorporates the Sankey and heat load diagrams to visualize energy flow and heat load accumulation across the cooling system. Results indicate that the secondary side yields the highest pressure drop, which emphasizes opportunities for hydraulic performance optimization. In addition, at the component level, the cooler’s low pressure drop highlights its energy efficiency while, at the system level, the cooler and tube connections make a significant contribution to system TUE, suggesting areas of improvement for energy efficiency. This framework relies on analytical methods without empirical validation and assumes steady-state conditions, which may limit its applicability. Future work should focus on experimental validation and transient thermal analysis. Additionally, the WFD methodology can be extended to different data center cooling methods and other thermal management systems such as hybrid air-liquid cooling, immersion cooling, HVAC, and battery cooling
Examining the social determinants of oral health using self-reported oral health measures and salivary biomarkers
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Kennedy Campbell, accepted the attached license on 2025-06-26 at 14:15.The student, Kennedy Campbell, submitted this Thesis for approval on 2025-06-26 at 16:44.This Thesis was approved for publication on 2025-06-30 at 10:07.DSpace SAF Submission Ingestion Package generated from Vireo submission #22372 on 2025-10-21 at 10:05:35Oral health is an important public health issue, as oral diseases affect millions of Americans. The purpose of the study is to evaluate the association between self-reported oral health, salivary biomarkers, and the social determinants of health (SDOH). This study uses a fully remote design with a non-invasive approach to assess oral health. Evaluating the effects of the social determinants of health is crucial to understanding the non-medical factors that may influence oral health outcomes, especially for marginalized communities affected by health disparities. This study utilized a cross-sectional design with 52 healthy adult participants across the United States. Participants provided saliva samples on three days and completed self-reported questionnaires. Biomarkers examined in this study included salivary IL-1β, IL-6, C-Reactive Protein, IgG, MMP-8, and transferrin concentrations. Key results of this study showed that MMP-8 had the strongest associations with self-reported teeth and gum health with higher concentrations associated with worse health, and IL-1β, IgG, and transferrin showed similar moderate associations. There were more significant associations between the SDOH and the self-reported oral health measures than the biomarker concentrations. Perceived stress, everyday discrimination, and air quality were statistically significantly associated with self-reported oral health. Education was statistically significant with self-reported teeth problems. IL-6 was significantly associated with neighborhood deprivation. This study provides a scalable approach for comprehensive investigations of oral health with future practice and policy implications
Regulation of rho guanine nucleotide exchange factors through lipid binding and phosphorylation
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Jesus Moreno Castillo, accepted the attached license on 2025-07-07 at 09:48.The student, Jesus Moreno Castillo, submitted this Dissertation for approval on 2025-07-07 at 10:47.This Dissertation was approved for publication on 2025-07-09 at 15:22.DSpace SAF Submission Ingestion Package generated from Vireo submission #22408 on 2025-10-21 at 10:05:38Rho guanine nucleotide exchange factors (RhoGEFs) are key activators of RhoGTPases, which are molecular switches that control the majority of cellular processes. The RhoGEF family consists of 71 members that activate 22 RhoGTPases. This disparity in the number of activators compared to substrates between RhoGEFs and RhoGTPases is believed to result from adaptation to signaling complexity during the evolution of multicellular organisms and an increase in cell-type diversity. Biochemically, this adaptation is highlighted by the domain diversity among the different members of the RhoGEF family. The different domains of RhoGEFs can determine the mechanism by which their function is regulated. RhoGEFs can be regulated through autoinhibition, protein-protein interaction, phosphorylation, and lipid binding. Phosphorylation and lipid binding are mechanisms that have been poorly elucidated in RhoGEFs. A single domain determines the activity of RhoGEFs, the Dbl-homology domain (DH). The Pleckstrin Homology (PH) domain is present in 67 of the 71 RhoGEFs and serves as a major regulator of the DH domain's activity. Although present in the majority of RhoGEFs, the PH domain exhibits low sequence similarity and contributes to the specific regulation of RhoGEF members. Additionally, the PH domain is known to bind to phosphatidylinositol phosphates (PIPs), which are signaling lipids that control virtually all cellular processes. Recently, the number of RhoGEFs that bind to PIPs has been expanded. However, the regulatory role of their interaction remains unknown. Similarly, PH domain phosphorylation has been reported in proteomic studies; however, its role remains uncharacterized for most RhoGEFs. Here, I investigate the role of phosphorylation and PIP binding of two RhoGEFs. In chapter 2, I elucidate the PH domain phosphorylation of ARHGEF3 by the novel type of PKCs. PKCs phosphorylate ARHGEF3 in the N-terminus and on S399, which is in the β5-β6 loop of the PH domain. I found that S399 phosphorylation impairs ARHGEF3 activation of RhoA in cells and in vitro. Additionally, we found that pS399 specifically disrupts ARHGEF3 binding to PI(3,5)P2 but not PI(4,5)P2. Next, we characterized a mutant that disrupts specifically PI(3,5)P2 and found that PI(3,5)P2 binding doesn’t affect ARHGEF3 activation of RhoA. Lastly, we found that S399 phosphorylation inhibited the intrinsic exchange activity of ARHGEF3. In chapter 3, I examine the regulation of the Rac1 GEF ARHGEF39 by PIP3 binding. I found that basic amino acids in the PH domain of ARHGEF39 regulate the interaction with PIP3. Additionally, the PH domain regulates ARHGEF39's binding to phosphatidylserine and its association with the plasma membrane. Next, we found that membrane association is sufficient for Rac1 activation. Structural analysis of the DH domain of ARHGEF39 suggests that it may be inactive. Lastly, protein-protein interaction predictions suggest that ARHGEF39 binds to Rac1-GTP, indicating an exchange-independent mechanism of Rac1 activation, which may be regulated by PIP3. In chapter 4, I summarize our findings and contrast them with the RhoGEF family. Through structural analysis and data mining, I found that the phosphorylation of RhoGEFs' PH domain is more prevalent than previously thought, opening the possibility of phosphorylation-dependent regulation of PIP binding as a common feature of RhoGEFs
Discovery and characterization of a set of peptide-modifying heme oxygenase-like domain-containing oxidases
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Alexander Battiste, accepted the attached license on 2025-07-09 at 10:11.The student, Alexander Battiste, submitted this Dissertation for approval on 2025-07-09 at 15:54.This Dissertation was approved for publication on 2025-07-17 at 11:10.DSpace SAF Submission Ingestion Package generated from Vireo submission #22456 on 2025-10-21 at 10:05:41Ribosomally synthesized and post-translationally modified peptides (RiPPs) are a class of natural products made through the post-translational modification of genetically encoded and ribosomally synthesized precursor peptides. Because of their wide array of bioactivities, it would be desirable to be able to catalog all RiPPs, both in terms of their structure and their bioactivities. However, while bioinformatic methods for RiPP identification have improved dramatically in the past decade, it is still challenging to identify RiPPs from new classes. This is especially true if the new RiPP class does not rely on enzymes that are homologous to enzymes from known classes of RiPPs. Even if a RiPP biosynthetic gene cluster (BGC) can be identified, there are often multiple residues that could possibly be modified by the enzymes, and it is not clear which of them will get modified. To determine the structure, and to determine the bioactivity of the natural product, the product must be purified, which can be a time consuming process. Two methods are described here, one method for the rapid assembly of RiPP BGCs into expression plasmids to facilitate their isolation from heterologous hosts, and one bioinformatic method for the discovery of new classes of RiPPs. Using these methods, a number of RiPPs from known classes are characterized, including some with bioactivities against known pathogens, and a new class of RiPPs is discovered. These new RiPPs are modified by a set of heme oxygenase domain-containing oxygenases (HDOs), an emerging class of metal dependent enzymes
Using environmental DNA to inform biodiversity conservation in agricultural landscapes
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Olivia Reves, accepted the attached license on 2025-07-09 at 21:35.The student, Olivia Reves, submitted this Thesis for approval on 2025-07-09 at 22:21.This Thesis was approved for publication on 2025-07-14 at 11:54.DSpace SAF Submission Ingestion Package generated from Vireo submission #22477 on 2025-10-21 at 10:05:45The conversion of natural ecosystems to agriculture is a leading cause of habitat loss and threatens global biodiversity. For the past two centuries, the Midwestern United States (U.S.) has experienced agricultural intensification and expansion, resulting in losses of native tallgrass prairies, wetlands, and forests. To mitigate agricultural pressures, conservation practices have been widely implemented across the Midwest. As a result, forest cover in Midwestern states like Illinois has increased over the last several decades partially due to agricultural conservation practices like riparian buffers, which can improve water quality, alleviate nutrient transfer and reduce soil loss. While the benefits of riparian buffers to aquatic taxa are well-documented, benefits to terrestrial biodiversity are less understood. To monitor wildlife across taxa, a combination of conventional methods can be costly, time-consuming and invasive. Environmental DNA (eDNA), DNA collected and isolated from environmental samples, can provide a solution to this challenge. Further, rivers have proven to be conveyor belts of not only aquatic DNA but also terrestrial eDNA as well from upstream or adjacent riparian areas. Among many factors that can influence eDNA dynamics in a river, precipitation can either dilute eDNA due to increasing discharge or mobilize eDNA into rivers from adjacent riparian terrestrial ecosystems. For my first study, I sampled three rivers before and after precipitation events in the Vermilion River watershed of east-central Illinois to evaluate if terrestrial DNA exhibits a mobilization effect and if aquatic eDNA exhibits a dilution effect in response to rainfall. I found that rainfall had a positive, significant effect on terrestrial taxa richness and a negative, non-significant effect on aquatic taxa richness. This study not only advances eDNA applications for vertebrate community analysis but also informs future eDNA sampling efforts for detecting terrestrial and aquatic taxa from lotic environments. For my second study, I collected eDNA samples from streams across a forest cover gradient in central Illinois to quantify the terrestrial biodiversity co-benefit of riparian buffers, a proposed agricultural conservation practice for the Midwest, to terrestrial wildlife. I demonstrated that sites with complete (100%) riparian buffers can support three times the terrestrial vertebrate taxa richness compared to sites lacking buffers entirely, with highly forested sites harboring unique forest-associated communities. This study demonstrates the benefit of riparian buffers to terrestrial wildlife and validates eDNA as a rapid biomonitoring tool to assess biodiversity responses to agricultural conservation practices. Collectively, my thesis research seeks to support biodiversity conservation across agricultural landscapes
Innovations in imaging sequences and reconstruction pipelines for ultrasound localization microscopy
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, YiRang Shin, accepted the attached license on 2025-07-11 at 11:42.The student, YiRang Shin, submitted this Dissertation for approval on 2025-07-11 at 11:46.This Dissertation was approved for publication on 2025-07-17 at 06:03.DSpace SAF Submission Ingestion Package generated from Vireo submission #22491 on 2025-10-21 at 10:05:47The microcirculation network, consisting of arterioles, capillaries, and venules, is the primary site for exchanging oxygen, nutrients, and waste products. This network is crucial for maintaining tissue viability and function. Imaging the microvasculature is vital because it reveals subtle changes in tissue perfusion and vessel integrity -often precursors to overt organ dysfunction. Such imaging enables earlier disease detection, targeted therapies, and monitoring treatment outcomes across various pathologies. In the past decade, Ultrasound localization microscopy (ULM) has emerged as a super-resolution imaging technique capable of imaging microvascular networks in deep tissue. It achieves this by localizing and tracking intravascular microbubbles at sub-wavelength precision, bypassing the resolution-penetration trade-off of ultrasound. ULM provides micron-scale spatial resolution at a depth unattainable by optical modalities and offers finer in vivo vascular detail than that of MRI and CT. ULM uses widely accessible ultrasound systems and FDA-approved microbubbles, giving it great promise for clinical use. It can also complement conventional vascular imaging by offering a unique balance of resolution, depth, and functional insight. However, translating ULM into clinical practice faces several challenges. A primary limitation is the prolonged data acquisition time, as it requires accumulating thousands of frames to capture a sufficient microbubble localization event. The technique also relies on a serial post-processing pipeline (beamforming, clutter filtering, localization, tracking) that limits its real-time application. Furthermore, ULM has inherent limitations stemming from 2D planar imaging, which cannot accurately capture complex 3D vascular architectures. To address these current ULM limitations, this thesis introduces a comprehensive framework incorporating innovative imaging sequences and advanced reconstruction pipelines for ULM. Specifically, we present LOCA-ULM, a deep learning-based simulation and localization pipeline that enables robust microbubble localization even at high concentrations, thereby substantially reducing acquisition time. Moreover, we adopt Trackformer to replace the computationally expensive microbubble detection and tracking stages with an end-to-end tracking-by-attention framework. Finally, we introduce Fast3D-AM, a high-volume acquisition rate 3D nonlinear imaging method using multiplexed 2D matrix arrays to reduce tissue clutter signals and facilitate robust, high-quality 3D ULM. We validate these frameworks with structural, functional, and volumetric imaging of diverse organs in small animal models and demonstrate its capability for comprehensive microvascular visualization and hemodynamic mapping. These technological advances represent a significant step towards the successful translation of ULM into both preclinical research and clinical settings
Understanding institutional agents’ engagement and function in supporting undocumented and DACA students in higher education across academic disciplines
No full textSubmission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2027-08-01The student, Hyun Jung Kim, accepted the attached license on 2025-07-14 at 13:07.The student, Hyun Jung Kim, submitted this Dissertation for approval on 2025-07-14 at 13:16.This Dissertation was approved for publication on 2025-07-16 at 15:54.DSpace SAF Submission Ingestion Package generated from Vireo submission #22550 on 2025-10-21 at 10:06:03Undocumented and Deferred Action for Childhood Arrivals (DACA) students face persistent legal, financial, and social barriers in higher education. Institutional agents—faculty, staff, and administrators—play a critical role in supporting these students through resource connection, service provision, policy implementation, and everyday interactions. However, institutional responses are often inconsistent due to agents’ limited awareness, lack of knowledge, insufficient guidance, and systemic constraints that hinder effective support. This qualitative study examines how institutional agents at a large public research university in Illinois perceive and enact their roles in supporting undocumented and DACA students. It explores four interrelated dimensions: (1) awareness of undocumented and DACA students’ presence and challenges; (2) knowledge of needs, policies, and resources; (3) perspectives on roles and responsibilities related to support practices; and (4) variation across academic disciplines and institutional positions. Through these dimensions, the study also reveals the challenges agents face and the strategies they recommend for strengthening institutional capacity to support undocumented and DACA students more effectively. Using stratified purposeful sampling, nineteen institutional agents, including faculty, staff, and administrators from ten of the fifteen colleges at a large Midwest public university, were interviewed in Fall 2024. Data were analyzed using a hybrid coding strategy that combined inductive thematic analysis with theoretical constructs from Bronfenbrenner’s Ecological Systems Framework, social and cultural capital theories, Bandura’s Social Cognitive Theory, and Diffusion of Responsibility Theory. Agents’ awareness of undocumented and DACA students and their challenges ranged from low-average to high-average, regardless of professional title or academic discipline. Awareness was shaped more by agents’ personal background, direct experience with undocumented/DACA students, and participation in ally trainings than by their formal institutional role itself. Even highly experienced agents frequently described their awareness as “average,” underscoring the evolving nature of competence based on diverse degrees of experience with undocumented and DACA students. Trained agents exhibited deeper knowledge of relevant policies and institutional resources, enabling them to implement more inclusive and adaptive strategies. In contrast, untrained agents often held fragmented understandings and felt underprepared, frequently referring students to diversity or cultural centers without offering further assistance. All participants expressed a strong sense of moral and professional responsibility for supporting students’ academic, emotional, social, and professional well-being. While faculty tended to focus on academic support, staff emphasized advising and student services, and administrators concentrated on policy, programming, and leadership. Yet agents across all roles recognized that undocumented/DACA students’ challenges are interconnected. Legal precarity, fear of deportation, and lack of access to opportunities often disrupt students’ educational trajectories, career planning, and emotional stability—making holistic, cross-functional approaches essential. Participants also noted that students often bear the burden of disclosing their immigration status to access tailored services—an expectation that can deter help-seeking and increase vulnerability. To reduce this burden and enhance institutional responsiveness, agents recommended establishing a centralized resource hub, designating personnel within departments to maintain current information, increasing the visibility of services, and implementing mandatory, role-specific, practice-oriented training. This support would not only benefit students but also equip agents, especially those in support roles or seeking to become more involved with the tools and confidence needed to engage more effectively. Additionally, participants underscored that trust-based relationships are foundational to meaningful support. They emphasized the importance of understanding students’ cultural backgrounds and lived experiences, ideally through listening directly to students’ voices, for culturally responsive engagement. While participants demonstrated deep commitment to equity, they also identified institutional barriers such as inconsistent messaging across departments, underutilized training opportunities, and decentralized structures as persistent obstacles. This study concludes by calling for institutional investments in centralized infrastructure, coordinated communication, and mandatory, ongoing ally training visibly supported by senior leadership to cultivate a university-wide culture of equity, inclusion, and informed engagement