Theses and Dissertations
Advisor
Chaudhary, Vini
Committee Member
Perkins, Andy
Committee Member
Manias, Dimitrios
Date of Degree
8-12-2026
Original embargo terms
Embargo 6 months
Document Type
Graduate Thesis - Open Access
Major
Cybersecurity & Operations
Degree Name
Master of Science (M.S.)
College
James Worth Bagley College of Engineering
Department
Department of Computer Science and Engineering
Abstract
Quantum communication networks require effective routing strategies to establish high-fidelity entanglement between distant nodes. Existing learning-based routing methods often assume fixed quantum memory characteristics, although memory coherence can vary over time due to hard-ware and environmental effects. This thesis studies route selection under time-varying memory conditions by modeling memory coherence as a temporally correlated stochastic process. The routing problem is formulated as a multi-armed bandit task in which paths are sequentially selected and evaluated through simulated network rewards derived from end-to-end fidelity. A standard Upper Confidence Bound method is compared with a sliding-window variant that emphasizes recent observations. Simulation results show that the standard method provides stable long-term reward performance but often concentrates on a dominant route, while the sliding-window method maintains more balanced path exploration with comparable reward performance. These findings demonstrate the value of adaptive learning strategies for routing in time-varying quantum network environments.
Recommended Citation
Joseph, Anthony, "Learning-based routing in quantum communication networks under time-varying memory decoherence" (2026). Theses and Dissertations. 7025.
https://scholarsjunction.msstate.edu/td/7025
