Defense Notices
All students and faculty are welcome to attend the final defense of EECS graduate students completing their M.S. or Ph.D. degrees. Defense notices for M.S./Ph.D. presentations for this year and several previous years are listed below in reverse chronological order.
Students who are nearing the completion of their M.S./Ph.D. research should schedule their final defenses through the EECS graduate office at least THREE WEEKS PRIOR to their presentation date so that there is time to complete the degree requirements check, and post the presentation announcement online.
Upcoming Defense Notices
Muhammad Momin Rahman
Development of a Signal Flow Graph-Based De-Embedding Framework for Broadband Passive Linear Multiport RF & Microwave Networks with Applications to Power SplittersWhen & Where:
Nichols Hall, Room 317 (Richard K. Moore Conference Room)
Committee Members:
Carl Leuschen, ChairKenneth Demarest
Fernando Rodriguez-Morales
Abstract
Discrete passive components, including resistors, capacitors, inductors, and transformers, are fundamental to broadband 75-Ω radio-frequency (RF) systems, particularly in cable television (CATV) and hybrid fiber-coaxial (HFC) networks. These components are incorporated into power splitters, couplers, power inserters, and amplifiers to achieve desired frequency responses, impedance matching, and signal transmission characteristics. However, their electrical behavior becomes increasingly complex at higher frequencies due to parasitic resistance, inductance, capacitance, electromagnetic coupling, and distributed transmission-line effects. Consequently, accurate broadband characterization is essential for predicting component behavior and developing representative RF circuit models for simulation purposes.
A significant challenge arises when characterizing passive components with limited or unavailable manufacturer specifications, particularly when existing component inventories must be evaluated for broadband applications. Conventional measurement instruments, including digital multimeters, LCR meters, oscilloscopes, and function generators, generally lack the frequency range required for comprehensive RF characterization from 5 MHz to 3 GHz. Vector network analyzers (VNAs) enable complex scattering-parameter measurements across this frequency range; however, the measured response includes contributions from test fixtures, interconnections, and parasitic discontinuities that must be distinguished from the intrinsic response of the device under test (DUT).
This project investigates the development of a signal flow graph (SFG)-based de-embedding framework for broadband passive linear multiport RF networks. The methodology incorporates series-through and shunt-through DUT (CPWG) fixture configurations, conventional Open-Short and Short-Open de-embedding techniques, followed by lumped-element and distributed transmission-line representations of fixture parasitics. Scattering-parameter and ABCD-parameter analysis are employed to characterize fixture responses, while signal flow graph derivations provide a mathematical basis for analyzing interconnected multiport networks. Particular attention is given to fixture asymmetry, impedance discontinuities, and multiple signal propagation paths that complicate the extension of conventional two-port characterization methods to higher-order networks.
Experimental validation is conducted through the characterization of passive components with published impedance and quality-factor characteristics, enabling comparisons between extracted measurements, equivalent-circuit models, and manufacturer reference data. The methodology is subsequently applied to broadband power splitter networks used in CATV/HFC systems to investigate the accuracy and applicability of the proposed multiport de-embedding framework. The project aims to establish a systematic approach to broadband passive component characterization that integrates fixture modeling, network theory, and experimental measurement while reducing dependence on proprietary characterization data.
Arya Hadizadeh Moghaddam
Learning Personalized and Robust Patient Representations across Complex Structures in Electronic Health RecordsWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
Zijun Yao, ChairFengjun Li
Bo Luo
Han Wang
Xinmai Yang
Abstract
Electronic health records (EHRs) contain medical codes and clinical notes collected over multiple visits. This dissertation studies how to learn patient representations from these records through three related aspects problems:
The first aspect concerns relationships within a patient’s history. TPGT models changing dependencies among medical codes within and across visits, and ARCI separates temporal patterns associated with different medical intents. They account for interacting clinical events and concurrent treatment goals.
When patient histories are limited, these patterns are harder to learn. The second aspect is addressed by MetaDrug, which adapts to patients with sparse histories using their own records and those of similar patients, while filtering uncertain samples.
The third aspect concerns incorporating clinical notes and large language models into patient representations. RePrompT uses cohort-trained representations and recurrent soft prompts to preserve longitudinal context. RASPER trains a note summarizer through downstream prediction feedback. CERE reasons over notes conditioned on structured codes to represent the additional evidence they provide. These studies connect structured histories with clinical text to support prediction.
Experiments on the KUMC acute kidney injury cohort, eICU, MIMIC-III, and MIMIC-IV show improvements over the respective baselines in clinical prediction and medication recommendation.
Shelby Jones
Measuring LLM Susceptibility to Stylistic Perturbations in Social Media TextWhen & Where:
Nichols Hall, Room 246 (Executive Conference Room)
Committee Members:
Jennifer Lohoefener, ChairPerry Alexander
John Symons
Abstract
Large Language Models (LLMs) are becoming increasingly integrated into important domains such as social media and politics. This makes it essential that we understand their susceptibility to superficial text variations. In this thesis, I investigated whether small stylistic perturbations in social media posts influence how an LLM perceives a post's political leanings. To do this, I systematically manipulated punctuation and capitalization across both political and apolitical social media posts. I then used VADER sentiment analysis to measure the changes in semantic intensity to quantify the text manipulation. Finally, I prompted each LLM to provide its perceived political leanings of each given post. I used linear regression to assess the VADER score correlation with the LLMs' political alignment ratings. While there was no statistically significant relation between perceived political extremity and VADER scores, category-level analysis revealed that the stylistic perturbations cause significant inconsistencies in model outputs. Furthermore, in posts with neutral semantic content, the stylistic perturbations introduce a right-leaning bias. These findings add to the growing body of literature on LLM instability under prompt perturbation and highlight risks associated with automated content analysis.
Connor Sullivan
Memory QoS Mechanisms for Predictable and Performant Real-Time SystemsWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
Heechul Yun, ChairZitao Chen
Tamzidul Hoque
Prasad Kulkarni
Kenneth Fischer
Abstract
Commercial-off-the-shelf (COTS) multicore processors are increasingly used in safety-critical real-time systems due to their higher performance and improved size, weight, and power (SWaP) characteristics. However, modern platforms are designed for average-case performance, which works in direct opposition to the worst-case timing guarantees required by safety-critical systems.
In particular, unregulated sharing of the memory hierarchy across cores can cause severe and unpredictable interference. When multiple cores concurrently access shared memory resources, their requests contend with one another, degrading application performance and undermining timing guarantees. Hardware quality-of-service (QoS) mechanisms aim to manage this contention. Partitioning these resources in space alone does not guarantee temporal isolation. Bandwidth regulation can protect latency-critical tasks from such interference, but it trades system throughput for isolation. Current solutions throttle co-running tasks far more than necessary, offering system designers a poor isolation-throughput tradeoff.
Addressing this challenge, this thesis focuses on QoS mechanisms that improve the balance between throughput and isolation. Shared memory resources such as the last-level cache (LLC) and main memory (DRAM) have parallel, banked structures, yet existing regulation treats them as monolithic. This thesis shows that contention in both arises at the bank level and proposes per-bank bandwidth regulation for the LLC and DRAM, which preserves isolation for latency-critical tasks while recovering throughput for best-effort tasks. Finally, the thesis aims to coordinate regulation across the levels of the memory hierarchy, adapting it dynamically at runtime. It will also explore matching the control mechanism to the root cause of contention, for example by controlling the number of outstanding requests, as an alternative or cooperative means to bandwidth regulation. Together, these efforts work toward a better balance between system throughput and isolation guarantees in multicore real-time systems.
Lee Taylor
Ultrawideband Single-Pass Interferometric SAR Integrated with Multi-Rotor UAVWhen & Where:
Nichols Hall, Room 317 (Richard K. Moore Conference Room)
Committee Members:
Carl Leuschen, ChairShannon Blunt
Patrick McCormick
John Paden
Fernando Rodriguez-Morales
Abstract
Ultrawideband (UWB) Interferometric Synthetic Aperture Radar (InSAR) integrated with multi-rotor Uncrewed Aerial Vehicle (UAV), or UIMU in this work for brevity, provides ultrafine-resolution, all-weather, three-dimensional (3D) surface imagery at any time of day. UIMU can be rapidly deployed and is low-cost, and therefore a critical new tool for low-altitude remote sensing applications, such as disaster response, environmental monitoring, and intelligence, surveillance, and reconnaissance (ISR). Traditional repeat-pass data collection methods reduce the phase coherence required for InSAR processing of ultrafine-resolution datasets due to the unstable flight behavior of multi-rotor UAVs. Collecting Synthetic Aperture Radar (SAR) datasets using two receive channels during a single pass improves phase coherence and the ability to produce ultrafine-resolution 3D InSAR imagery.
This work quantifies and characterizes the 3D target-position accuracy of a dual-channel 6 GHz bandwidth (2.5 cm slant range resolution) frequency modulated continuous wave (FMCW) radar integrated with the Aurelia X6 hexacopter, establishing novel single-pass UWB InSAR data collection methods and processing algorithms for multi-rotor UAVs. Building on earlier qualitative single-pass InSAR imagery that demonstrated the feasibility of the approach, fieldwork was conducted to measure the positions of Global Navigation Satellite System (GNSS) located corner reflectors using the UIMU system. Algorithms for motion tolerant Time-Domain Backprojection (TDBP), InSAR coregistration, and digital elevation mapping, novel to multi-rotor UAVs at UWB, were developed and are presented, and the TDBP imagery is compared with conventional frequency-domain Omega-K processing. The effect of vehicle motion on the synthetic aperture is investigated. The results show that UIMU is possible, with calibration of the signal processing workflow being of utmost importance: the radar, the GNSS logger and the flight log must be aligned closely in time, because the position and attitude of the vehicle vary rapidly over short time scales and coherence is quickly lost as the position or alignment error grows. The internal delay and phase of the radar must also be precisely calibrated. With these calibrations, single-pass interferometry from the multi-rotor UAV is achieved: both corner reflectors focus over a 27 m synthetic aperture, and the interferometrically detected height difference between the two reflectors, 0.28 ± 0.08 m, falls within 3 cm of their surveyed 0.25 m height offset. The radar-measured reflector positions differ from the in-situ real-time kinematic (RTK) GNSS positions by about 25 cm in ground range, at a ground range resolution of about 3.5 cm; this difference is largely attributed to sensitivity to calibration and to the precision of the radar antenna positions. Using one RTK reference point for the ground height, the radar-measured ground elevations agree with the in-situ RTK measurements to within 0.20 m rms and lie 0.2 to 0.6 m above the TessaDEM global elevation model. Centimeter-scale error sources behind a 6.5 cm residual slant range offset and the remaining loss of focus are investigated, estimating how sensitive UIMU is to calibration errors, to errors in the lever arms to the antenna phase centers, and to errors in the measured vehicle position and attitude, which shift and defocus the image.
This work establishes a foundation for future investigations of real-time image processing, separated transmit and receive platforms (bistatic), and swarm configurations for UIMU systems.
Past Defense Notices
Zhaohui Wang
Detection and Mitigation of Cross-App Privacy Leakage and Interaction Threats in IoT AutomationWhen & Where:
Nichols Hall, Room 250 (Gemini Conference Room)
Committee Members:
Fengjun Li, ChairAlex Bardas
Drew Davidson
Bo Luo
Haiyang Chao
Abstract
The rapid growth of Internet of Things (IoT) technology has brought unprecedented convenience to everyday life, enabling users to deploy automation rules and develop IoT apps tailored to their specific needs. However, modern IoT ecosystems consist of numerous devices, applications, and platforms that interact continuously. As a result, users are increasingly exposed to complex and subtle security and privacy risks that are difficult to fully comprehend. Even interactions among seemingly harmless apps can introduce unforeseen security and privacy threats. In addition, violations of memory integrity can undermine the security guarantees on which IoT apps rely.
The first approach investigates hidden cross-app privacy leakage risks in IoT apps. These risks arise from cross-app interaction chains 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 risk levels based on inferences. In addition, we provide a fine-grained categorization of privacy threats to generate detailed alerts, enabling users to better understand and address specific privacy risks.
The second approach addresses cross-app interaction threats in IoT automation systems by leveraging a logic-based analysis model grounded in event relations. We formalize event relationships, detect event interferences, and classify rule conflicts, then generate risk scores and conflict rankings to enable comprehensive conflict detection and risk assessment. To mitigate the identified interaction threats, an optimization-based approach is employed to reduce risks while preserving system functionality. This approach ensures comprehensive coverage of cross-app interaction threats and provides a robust solution for detecting and resolving rule conflicts in IoT environments.
To support the development and rigorous evaluation of these security analyses, we further developed a large-scale, manually verified, and comprehensive dataset of real-world IoT apps. This clean and diverse benchmark dataset supports the development and validation of IoT security and privacy solutions. All proposed approaches are evaluated using this dataset of real-world apps, collectively offering valuable insights and practical tools for enhancing IoT security and privacy against cross-app threats. Furthermore, we examine the integrity of the execution environment that supports IoT apps. We show that, even under non-privileged execution, carefully crafted memory access patterns can induce bit flips in physical memory, allowing attackers to corrupt data and compromise system integrity without requiring elevated privileges.
Shawn Robertson
A Low-Power Low-Throughput Communications Solution for At-Risk Populations in Resource Constrained Contested EnvironmentsWhen & Where:
Nichols Hall, Room 246 (Executive Conference Room)
Committee Members:
Alex Bardas, ChairDrew Davidson
Fengjun Li
Bo Luo
Shawn Keshmiri
Abstract
In resource‑constrained contested environments (RCCEs), communications are routinely censored, surveilled, or disrupted by nation‑state adversaries, leaving at‑risk populations—including protesters, dissidents, disaster‑affected communities, and military units—without secure connectivity. This dissertation introduces MeshBLanket, a Bluetooth Mesh‑based framework designed for low‑power, low‑throughput messaging with minimal electromagnetic spectrum exposure. Built on commercial off‑the‑shelf hardware, MeshBLanket extends the Bluetooth Mesh specification with automated provisioning and network‑wide key refresh to enhance scalability and resilience.
We evaluated MeshBLanket through field experimentation (range, throughput, battery life, and security enhancements) and qualitative interviews with ten senior U.S. Army communications experts. Thematic analysis revealed priorities of availability, EMS footprint reduction, and simplicity of use, alongside adoption challenges and institutional skepticism. Results demonstrate that MeshBLanket maintains secure messaging under load, supports autonomous key refresh, and offers operational relevance at the forward edge of battlefields.
Beyond military contexts, parallels with protest environments highlight MeshBLanket’s broader applicability for civilian populations facing censorship and surveillance. By unifying technical experimentation with expert perspectives, this work contributes a proof‑of‑concept communications architecture that advances secure, resilient, and user‑centric connectivity in environments where traditional infrastructure is compromised or weaponized.
Shravan Kaundinya
Design, Development, And Deployment of Airborne and Ground-Based High-Power, UHF Radars With Multichannel, Polarimetric Antenna Arrays for RadioglaciologyWhen & Where:
Nichols Hall, Room 317 (Moore Conference Room)
Committee Members:
Carl Leuschen, ChairRachel Jarvis
John Paden
Jim Stiles
Richard Hale
Abstract
This work describes the building and deployment of airborne and ground-based high-power, UHF radars from a systems engineering perspective. Its primary focus is on the design and development of compact, low-profile, polarimetric antenna arrays for these radars using a rapid prototyping methodology. The overarching goal of this effort is to aid the Center for Oldest Ice Exploration (COLDEX), a multi-institution collaboration to explore Antarctica using airborne and ground radars for the identification of a drill site to retrieve the oldest possible continuous ice record. A multichannel 600 – 900 MHz, pulsed frequency modulated (FM) radar with up to 1.6 kW of peak output power per channel is designed and implemented. The ground-based frontend is a 16-element antenna array power-combined into a single channel per polarization in a sled platform. The airborne frontend has a 64-element fuselage-mounted antenna array power-combined into 16 independent channels and two 12-element wing arrays power-combined into 6 channels for operation on a Basler aircraft.
Three major design revisions of the antenna element design are presented. The first two design revisions of the dual-polarized, microstrip dipole antenna have the typical vertically integrated aperture-coupled microstrip baluns. The third and newly proposed design is a near-planar, integrated feed which combines a 2-sided microstrip balun board (one balun for each polarization) and a custom 6-layer balanced-to-balanced feed board. A microstrip matching network 2-layer board with two order-4 LC-filters is directly connected using micro-coaxial (MCX) connectors. The total antenna height of the proposed design is reduced by nearly one-third relative to the first two design revisions while improving electrical performance.
A novel methodology for efficient wideband tuning of the active impedance of the elements of an antenna array using lumped components is demonstrated. The goal of the method is to achieve >10 dB active return loss with a single order-4 LC-circuit for all four power-combined channels of the 16-element antenna array with minimal iteration loops. It combines the simulation and measurement spaces at different stages to account for platform scattering, mutual coupling, and non-ideal behavior of the lumped components and circuit board parasitic effects in the UHF range.
Each antenna array design is fed using 1:2 and 1:4 microstrip, Wilkinson high-power dividers. Two major design revisions of the high-power divider are presented. The first design has three implementations: ground-based, airborne fuselage-mounted, and airborne wing-mounted. It uses a 100-ohm flange resistor under the requirements of fire safety in the case of all transmitted power reflected from the antenna port. Two drawbacks of the flange design feature are high parasitic capacitance (which results in sub-optimal performance) and large profile. The second and newly proposed design uses chemical vapor deposition (CVD) diamond resistors on a custom copper flange. The resistors are wire-bonded between the resistor’s gold contacts and soft gold pads on the circuit board using 25 µm gold wire. Results for an ideal prototype and the first implemented version on a ground-based array are presented. System engineering aspects such as thermal cycling, high-power RF tests, and bond integrity are explored.
The effectiveness of the circuits developed in the context of this work is demonstrated in real field environments. This includes the operation of the airborne version of the UHF multichannel radar for surveys near Dome A in Antarctica during the 2022 – 2023 and 2023 – 2024 Austral summer seasons, the five-fold deployment of the ground-based versions of the UHF multielement radar for surveys in Greenland and Antarctica from 2022 to 2024, and the operation of the newly proposed version to Taylor Dome in Antarctica during the 2025 Austral summer season, currently underway.
Sai Karthik Maddirala
Real-Estate Price Analysis and Prediction Using Ensemble LearningWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
David Johnson, ChairMorteza Hashemi
Prasad Kulkarni
Abstract
Accurate real-estate price estimation is crucial for buyers, sellers, investors, lenders, and policymakers, yet traditional valuation practices often rely on subjective judgment, inconsistent expertise, and incomplete market information. With the increasing availability of digital property listings, large volumes of structured real-estate data can now be leveraged to build objective, data-driven valuation systems. This project develops a comprehensive analytical framework for predicting different types of properties prices using real-world listing data collected from 99acres.com across major Indian cities. The workflow includes automated web scraping, extensive data cleaning, normalization of heterogeneous property attributes, and exploratory data analysis to identify important pricing patterns and structural trends within the dataset. A multi-stage learning pipeline is designed—consisting of feature preparation, hyperparameter tuning, cross-validation, and performance evaluation—to ensure that the final predictive system is both reliable and generalizable. In addition to the core prediction engine, the project proposes a future extension using Retrieval-Augmented Generation (RAG) with Large Language Models(LLM’s) to provide transparent, context-aware explanations for each valuation. Overall, this work establishes the foundation for a scalable, interpretable, and data-centric real-estate valuation platform capable of supporting informed decision-making in diverse market contexts.
Ramya Harshitha Bolla
AI Academic Assistant for Summarization and Question AnsweringWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
David Johnson, ChairRachel Jarvis
Prasad Kulkarni
Abstract
The rapid expansion of academic literature has made efficient information extraction increasingly difficult for researchers, leading to substantial time spent manually summarizing documents and identifying key insights. This project presents an AI-powered Academic Assistant designed to streamline academic reading through multi-level summarization, contextual question answering, and source-grounded traceability. The system incorporates a robust preprocessing pipeline including text extraction, artifact removal, noise filtering, and section segmentation to prepare documents for accurate analysis. After assessing the limitations of traditional NLP and transformer-based summarization models, the project adopts a Large Language Model (LLM) approach using the Gemini API, enabling deeper semantic understanding, long-context processing, and flexible summarization. The assistant provides structured short, medium, and long summaries; contextual keyword extraction; and interactive question answering with transparent source highlighting. Limitations include handling complex visual content and occasional API constraints. Overall, this project demonstrates how modern LLMs, combined with tailored prompt engineering and structured preprocessing, can significantly enhance the academic document analysis workflow.
Keerthi Sudha Borra
Intellinotes – AI-POWERED DOCUMENT UNDERSTANDING PLATFORMWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
David Johnson, ChairPrasad Kulkarni
Han Wang
Abstract
This project presents Intellinotes, an AI-powered platform that transforms educational documents into multiple learning formats to address information-overload challenges in modern education. The system leverages large language models (GPT-4o-mini) to automatically generate four complementary outputs from a single document upload: educational summaries, conversational podcast scripts, hierarchical mind maps, and interactive flashcards.
The platform employs a three-tier architecture built with Next.js, FastAPI, and MongoDB, supporting multiple document formats (PDF, DOCX, PPTX, TXT, images) through a robust parsing pipeline. Comprehensive evaluation on 30 research documents demonstrates exceptional system reliability with a 100% feature success rate across 150 tests (5 features × 30 documents), and strong semantic understanding with a semantic similarity score of 0.72.
While ROUGE scores (ROUGE-1: 0.40, ROUGE-2: 0.09, ROUGE-L: 0.17) indicate moderate lexical overlap typical of abstractive summarization, the high semantic similarity demonstrates that the system effectively captures and conveys the conceptual meaning of source documents—an essential requirement for educational content. This validation of meaning preservation over word matching represents an important contribution to evaluating educational AI systems.
The system processes documents in approximately 65 seconds with perfect reliability, providing students with comprehensive multi-modal learning materials that cater to diverse learning styles. This work contributes to the growing field of AI-assisted education by demonstrating a practical application of large language models for automated educational content generation supported by validated quality metrics.
Sowmya Ambati
AI-Powered Question Paper GeneratorWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
David Johnson, ChairPrasad Kulkarni
Dongjie Wang
Abstract
Designing a well-balanced exam requires instructors to review extensive course materials, determine key concepts, and design questions that reflect appropriate difficulty and cognitive depth. This project develops an AI-powered Question Paper Generator that automates much of this process while keeping instructors in full control. The system accepts PDFs, Word documents, PPT slides, and text files, extracts their content, and builds a FAISS-based retrieval index using sentence-transformer embeddings. A large language model then generates multiple question types—MCQs, short answers, and true/false—guided by user-selected difficulty levels and Bloom’s Taxonomy distributions to ensure meaningful coverage. Each question is evaluated with a grounding score that measures how closely it aligns with the source material, improving transparency and reducing hallucination. A React frontend enables instructors to monitor progress, review questions, toggle answers, and export to PDF or Word, while an ASP.NET Core backend manages processing and metrics. The system reduces exam preparation time and enhances consistency across assessments.
George Steven Muvva
Automated Fake Content Detection Using TF-IDF-Based Machine Learning and LSTM-Driven Deep Learning ModelsWhen & Where:
Eaton Hall, Room 2001B
Committee Members:
David Johnson, ChairRachel Jarvis
Prasad Kulkarni
Abstract
The rapid spread of misinformation across online platforms has made automated fake news detection essential. This project develops and compares machine learning (SVM, Decision Tree) and deep learning (LSTM) models to classify news headlines from the GossipCop and PolitiFact datasets as real or fake. After extensive preprocessing— including text cleaning, lemmatization, TF-IDF vectorization, and sequence tokenization—the models are trained and evaluated using standard performance metrics. Results show that SVM provides a strong baseline, but the LSTM model achieves higher accuracy and F1-scores by capturing deeper semantic and contextual patterns in the text. The study highlights the challenges of domain variation and subtle linguistic cues, while demonstrating that context-aware deep learning methods offer superior capability for automated fake content detection.
Babak Badnava
Joint Communication and Computation for Emerging Applications in Next-Generation Wireless NetworksWhen & Where:
Nichols Hall, Room 246 (Executive Conference Room)
Committee Members:
Morteza Hashemi, ChairVictor Frost
Prasad Kulkarni
Taejoon Kim
Shawn Keshmiri
Abstract
Emerging applications in next-generation wireless networks, such as augmented and virtual reality (AR/VR) and autonomous vehicles, demand significant computational and communication resources at the network edge. This PhD research focuses on developing joint communication–computation solutions while incorporating various network-, application-, and user-imposed constraints. In the first thrust, we examine the problem of energy-constrained computation offloading to edge servers in a multi-user, multi-channel wireless network. To develop a decentralized offloading policy for each user, we model the problem as a partially observable Markov decision process (POMDP). Leveraging bandit learning methods, we introduce a decentralized task offloading solution in which edge users offload their computation tasks to nearby edge servers over selected communication channels.
The second thrust focuses on user-driven requirements for resource-intensive applications, specifically the Quality of Experience (QoE) in 2D and 3D video streaming. Given the unique characteristics of millimeter-wave (mmWave) networks, we develop a beam alignment and buffer-predictive multi-user scheduling algorithm for 2D video streaming applications. This algorithm balances the trade-off between beam alignment overhead and playback buffer levels for optimal resource allocation across multiple users. We then extend our investigation to develop a joint rate adaptation and computation distribution framework for 3D video streaming in mmWave-based VR systems. Numerical results using real-world mmWave traces and 3D video datasets demonstrate significant improvements in video quality, rebuffering time, and quality variations perceived by users.
Finally, we develop novel edge computing solutions for multi-layer immersive video processing systems. By exploring and exploiting the elastic nature of computation tasks in these systems, we propose a multi-agent reinforcement learning (MARL) framework that incorporates two learning-based methods: the centralized phasic policy gradient (CPPG) and the independent phasic policy gradient (IPPG). IPPG leverages shared information and model parameters to learn edge offloading policies; however, during execution, each user acts independently based only on its local state information. This decentralized execution reduces the communication and computation overhead of centralized decision-making and improves scalability. We leverage real-world 4G, 5G, and WiGig network traces, along with 3D video datasets, to investigate the performance trade-offs of CPPG and IPPG when applied to elastic task computing.
Sri Dakshayani Guntupalli
Customer Churn Prediction for Subscription-Based BusinessesWhen & Where:
LEEP2, Room 2420
Committee Members:
David Johnson, ChairRachel Jarvis
Prasad Kulkarni
Abstract
Customer churn is a critical challenge for subscription-based businesses, as it directly impacts revenue, profitability, and long-term customer loyalty. Because retaining existing customers is more cost-effective than acquiring new ones, accurate churn prediction is essential for sustainable growth. This work presents a machine learning based framework for predicting and analyzing customer churn, coupled with an interactive Streamlit web application that supports real time decision making. Using historical customer data that includes demographic attributes, usage behavior, transaction history, and engagement patterns, the system applies extensive data preprocessing and feature engineering to construct a modeling-ready dataset. Multiple models Logistic Regression, Random Forest, and XGBoost are trained and evaluated using the Scikit-Learn framework. Model performance is assessed with metrics such as accuracy, precision, recall, F1-score, and ROC-AUC to identify the most effective predictor of churn. The top performing models are serialized and deployed within a Streamlit interface that accepts individual customer inputs or batch data files to generate immediate churn predictions and summaries. Overall, this project demonstrates how machine learning can transform raw customer data into actionable business intelligence and provides a scalable approach to proactive customer retention management.