Announcements

Nitesh Potu - Master's Thesis for M.S. in Computer Science
Title: On Further Reduction of the Diameter of an Existing Interest Based Peer-to-Peer Network Architecture
Major Professor: Bidyut Gupta
Committee Members : Anusuman Bhattacharya, Xiaolan Huang
Date: October 31, 2025
Location: Engineering A Conference Room 309C
Time: 3:00 p.m.
Abstract: In this work, we have proposed a modified version of the existing Residue-class (RC) based peer-to peer (P2P) network architecture. The existing RC based architecture [10] has been considered because of its two main advantages viz. (1) all peers with the same interest (or possessing the same resource type) structurally form a group of diameter one, and (2) the group heads are connected in the form of a ring and the ring always remains connected even in the presence of any churn. The diameter of the ring with n group heads is n/2 and as n increases, the latency of the existing data look-up protocols increases as well. In this paper our objective is to reduce this diameter substantially in order to reduce data look-up latency drastically. To achieve this, we have used a mathematical model using which we show that all group heads will be connected directly to each other; that is, instead of forming a ring structure of diameter n/2, it will be a network of diameter one only. Therefore, the diameter of the peer-to-peer network becomes only 3 instead of (n/2 + 2) as in [10]. This is the shortest diameter possible in such networks. It will definitely make any data look-up protocol highly efficient from the viewpoint of low latency.

Dheyaa Al-Sallami – Ph.D. Dissertation for Ph.D. in Molecular Biology, Microbiology and Biochemistry (MBMB)
Title: Unraveling neurodegenerative pathways and auditory deficits in 5xfad mice: Insights into adsensory alterations
Major Professors: Shelley Tischkau
Committee Members: Leonard Rybak, Andrew Wilber, Daotai Nie, Raheem Al Aameri
Date: October 15, 2025
Location: Virtual
Link: https://siumed.webex.com/siumed/j.php?MTID=m465c1db348d59b25301e2eb20fdae9d8
Time: 9:00 a.m
Abstract: The first detection of amyloid β42 (Aβ42) in the cochlea of 5XFAD mice reveals a new peripheral manifestation of Alzheimer’s disease. Aβ42 progressively accumulates in spiral ganglion neurons, the stria vascularis, and synapses, driving synaptopathy, neurodegeneration, and hearing loss through oxidative stress, mitochondrial dysfunction, and inflammation. Treatment with epigallocatechin gallate (EGCG) protects against this damage by lowering Aβ42 levels, reducing apoptosis, and preserving neuronal health. Reduced Wave I amplitude in auditory brainstem recordings emerges as an early, sensitive, and non-invasive biomarker of Alzheimer’s disease, preceding threshold shifts and cognitive decline.

Maryam Alzahrani – Ph.D. Dissertation for Ph.D. in Mathematics
Title: Some explicit descriptions of characters stabilizing smooth complex representations of $GL(r,D)$ with a $p$-adic division algebra $D$
Major Professors: Kwangho Choiy
Committee Members: Michael Sullivan, Jianhong Xu, Amer AbuGhazaleh, Dubtavka Ban
Date: October 23, 2025
Location: Neckers 156
Time: 3:00 p.m
Abstract: For a positive integer $r$ and a central division algebra $D$ over a non-archimedean field $F$ of characteristic zero, we provide an explicit and precise description of (the number of) characters that stabilize irreducible smooth complex representations of $GL(r,D)$. The works of Gelbart-Knapp and Hiraga-Saito on L-packets for inner forms of $SL(n)$ over the $p$-adic field are being mainly applied, and we also explicitly discuss that there are some representations that have multiplicity that are different than $1$.

Rasheed Ibrahim - Ph.D. Dissertation for Ph.D. in Criminology and Criminal Justice
Title: Why People Cooperate with the Police in Nigeria: An Alternative View
Major Professor: Tammy Kochel
Committee Members: Raymund Narag, Julie Hibdon, Sujung Cho, Justice Tankebe
Date: November 4, 2025
Location: Applied Sciences Building C-220
Time: 11:00 a.m.
Abstract: This study investigates why Nigerians cooperate with the police in the post-EndSARS era, focusing on social, emotional, and rational influences. Using survey data from about 1,600 adult Nigerians, the study applied structural equation modeling to test the normative, instrumental, expressive, and rational-choice theoretical frameworks. Results show that police legitimacy did not significantly predict cooperation, while dull compulsion—compliance driven by fear or coercion—was a strong predictor. Trust in the police, familiarity, and perceived benefits encouraged cooperation, whereas fear of the police, police corruption, and prior victimization discouraged it. These findings suggest that Nigerians may cooperate more out of fear and pragmatic calculation than moral obligation, challenging Western-based models. The study recommends reforms that reduce coercive practices, build trust, and promote community engagement to foster genuine and voluntary cooperation with the police.

Arya Sanjar - Master's Thesis for M.S. in Forestry
Title: Survivorship in Virginia Opossums and the Management Implications for Burmese Pythons in Key Largo, Florida
Major Professor: Brent Pease
Committee Members: Eric Holzmueller, Michael Cove
Date: October 27, 2025
Location: Virtual
Link: https://teams.microsoft.com/l/meetup-join/19%3ameeting_MDU3YzZlYTgtNTllOC00MWU2LWFhMWEtNDBkZjRlN2YyNWFm%40thread.v2/0?context=%7b%22Tid%22%3a%22d57a98e7-744d-43f9-bc91-08de1ff3710d%22%2c%22Oid%22%3a%2286ba6861-5225-46b8-bba8-c3e19bb61b44%22%7d
Time: 10:00 a.m.
Abstract: Invasive predators such as the Burmese python (Python bivittatus) pose a substantial threat to native fauna, particularly within insular ecosystems like Key Largo, Florida, USA. This study employed a telemetry-based framework, using radio-collared Virginia opossums (Didelphis virginiana), to quantify survival patterns and estimate detection probabilities of python predation events. Between April 2022 and January 2025, 135 adult opossums were monitored using VHF collars across protected habitats within Key Largo. Annual survival was estimated using a staggered-entry Kaplan-Meier model, producing an annual survival rate of 0.17 (SE = 0.089) for females, and male survival was estimated at 0.04 (SE=0.032) at 10 months since no male lived 10 months after collaring. Cause-specific monthly survival analyses indicated python predation peaked during the late summer months of July through September, when temperatures were at their annual highest. No python predation events were recorded from January through April, corresponding to lower temperatures and the python breeding season. Bootstrap simulations (n = 1000 per 5% increment) demonstrated that opossum survival estimates stabilize around 35% sample coverage or 48 collars, and produce RMSE values of 0.022, indicating that 135 collars is likely representative of the population. A python-specific detection model revealed a 95% detection probability with 8–10 collars, whereas a theoretical model, assuming independent mortality risk, required 37 collars to reach equivalent power. The empirical model consistently exceeded theoretical expectations, likely reflecting temporal clustering in predation risk. These findings support the integration of radio-collared prey into invasive predator monitoring protocols, providing a scalable, cost-effective approach for early detection and removal of invasive predators. Moreover, this method may be broadly transferable to similar habitats and comparable cryptic predators, offering strategic benefits in wildlife management and conservation monitoring.

Bhargav Thota - Ph.D. Dissertation for Ph.D. in Computer Science
Title: Assessing The Impact Of Social Network On Cyber-Argumentation
Major Professor: Xiaoqing “Frank” Liu
Committee Members: Chun-Hsi Huang, Khaled Ahmed, Zhong Chen, Brajendra Panda,
Date: October 27, 2025
Location: Engineering A Room 309C
Time: 10:30 a.m.
Abstract: Argumentation is one of the basic activities people perform in our society. People often spend a lot of time arguing about social, cultural, economic, financial, political, and other issues. Existing cyber-argumentation platforms facilitate large-scale discussions among participants and play an important role in the analysis of collective opinions and collective decision-making. However, they do not provide social networking capabilities and are not designed to be used in social networking environments. Social networks in argumentation platforms can reveal the patterns, shifts, and true relationships between the users in a discourse that influence the collective decision-making process, outlier opinions and so on. To this end, this research seek to address 1) Create a prototype of mobile application integrating social networking capabilities into cyber-argumentation platforms. 2) Assess the impact of social networks on the formation and evolution of collective opinion in argumentation. 3) Assess the impact of social networks on the formation of outlier opinions in argumentation by retrospective analysis. The proposed models and algorithm have been implemented within our intelligent cyber-argumentation system with social networking. These models were evaluated using empirical argumentation data and simulated social networking data. The empirical evaluations demonstrate the effectiveness of these proposed models and algorithms in cyber-argumentation with social networking.

Jatin Rai - Master's Thesis for M.S. in Computer Science
Title: Optimized Multi-Path Routing for IOT Connectivity in Leo Satellite Networks: A Load-Balanced Approach Networks
Major Professor: Ansuman Bhattacharya
Committee Members: Bidyut Gupta, Koushik Sinha,
Date: October 20, 2025
Location: Engineering A Room 309C
Time: 3:00 p.m.
Abstract: As the Internet of Things (IoT) ecosystem expands at an unprecedented rate, ensuring reliable and low-latency communication becomes increasingly critical. The surge in connected devices and data traffic has outpaced the capabilities of conventional ground-based networks, especially in isolated or underserved areas where real-time responsiveness is essential. To bridge this gap, Low Earth Orbit (LEO) satellite constellations offer a compelling alternative by enabling fast and widespread connectivity. Despite their potential, current routing strategies—particularly those relying on single-path transmission—struggle with issues like network congestion, unstable links, and frequent changes in topology. To overcome these limitations, this study introduces a multi-path routing protocol specifically designed for LEO satellite environments. The protocol dynamically models the satellite network using Two-Line Element (TLE) orbital data and applies the Edmonds-Karp algorithm to discover multiple node-disjoint routes for data delivery. To further optimize traffic flow, the protocol incorporates a Least Common Multiple-based Routing (LCMR) technique that evenly distributes packets across available paths. This approach minimizes latency and enhances overall network performance. Tailored for IoT systems that demand consistent and rapid communication between distributed endpoints, the proposed solution demonstrates marked improvements in throughput, load distribution, and end-to-end delay in simulation tests—positioning it as a robust candidate for the emerging Internet of Space Things (IoST)