Over the Air Quantum Federated Learning

This project investigates the fundamental theoretical boundaries of deploying Quantum Federated Learning over next-generation wireless networks. The scaling of distributed quantum intelligence is currently bottlenecked by the short coherence times of quantum hardware. While analog Over-the-Air (OtA) aggregation solves this latency issue by physically superimposing radio waves for instant computation, it creates a severe mathematical conflict: OtA operates in flat, Euclidean space, whereas quantum circuit parameters reside on curved manifolds governed by complex Lie algebras.

This research will mathematically bound the non-linear structural distortion caused by raw analog channel noise and design novel, noise-resilient quantum ansätze to survive it. By merging Korea University’s expertise in Quantum Machine Learning and differential geometry with Memorial University of Newfoundland’s leading expertise in physical-layer wireless communications, this project will deliver a joint communication-computation framework. Ultimately, it provides the theoretical blueprint required to achieve zero-latency, industrial-scale quantum edge networks.

Faculty Supervisor:

Octavia Dobre

Student:

Partner:

Korea University

Discipline:

Computer science

Sector:

Quantum Science; Information and Communications Technology (ICT)

University:

Memorial University of Newfoundland

Program: