[CCoE Notice] Cullen College Dissertation Announcement (ECE)

[Hutchinson, Inez A]

[Dissertation Defense Announcement at the Cullen College of Engineering]
Satellite-Terrestrial Integration under Uncertainty Using Machine Learning and Distributionally Robust Optimization
Kai-Chu Tsai

April 9, 2024; 9:30 AM - 11:00 AM (CST)
Location: ECE Large Conference Room D N328
Zoom:  https://urldefense.com/v3/__https://uh-edu-cougarnet.zoom.us/j/99146061274__;!!LkSTlj0I!DS5MvwsPz5i-L6CFrYwMd87ME49ZOSTD8_qLHpshO8DXIdObbBTh4yWS5lUrnFqlcGvlwEMSoLMuYOVGRbm0Np5F_NU$ 


Committee Chair:

Zhu Han, Ph.D. | Ricardo Lent, Ph.D

Committee Members:
Fan Lei, Ph.D. | Miao Pan, Ph.D. | Hien Van Nguyen, Ph.D. | Li-Chun Wang, Ph.D.

Abstract

Technological advances have accelerated the development of terrestrial communication networks in recent years, providing broadband connectivity accessible worldwide. However, a substantial portion of the earth is still disconnected, mostly due to persistent economic and geographic obstacles. This gap underlines the essential importance of Integrated Satellite-Terrestrial Networks (ISTNs), which use Low Earth Orbit (LEO) satellite constellations like Starlink, Telesat, and OneWeb, to connect isolated areas. This dissertation investigates the potential of ISTNs to revolutionize global communication infrastructures, focusing on their capacity to provide enhanced connection, lower latency, and higher throughput, which are crucial for real-time applications and global broadband coverage. Furthermore, it explores the uncertainties within these networks, offering novel solutions for dependable and effective global communication.

In the first work, we explore the LEO satellite constellations, employing deep reinforcement learning (DRL) to optimize multi-commodity flow routing. The second work investigates the task delivery problem within ISTNs, utilizing DRL to minimize total routing delay in task transfer among user equipment. The last work concentrates on the uncertainty within ISTNs. To address these issues, we develop a distributionally robust optimization (DRO) model that minimizes the worst-case overall task routing latency under uncertain probability distributions. Utilizing the Wasserstein ambiguity set, the model effectively accommodates unidentified vehicle movement and intermittent connections, offering an optimal routing path for task uploading, satellite constellation, and task downloading. Collectively, these studies provide a complete approach to tackling complex present-day network routing and management challenges.

[Engineered For What's Next]

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://Bug.EGR.UH.EDU/pipermail/engi-dist/attachments/20240408/e4d78397/attachment-0001.html