 |
Intelligent Routing and Tomography for Future Quantum Networks (Xuchuang Wang et al.)
The development of quantum networks promises to revolutionize information technology by enabling unconditionally secure communication and distributed quantum computing. To realize the "Quantum Internet," advanced algorithms must be developed to manage the generation and distribution of quantum entanglement across complex topologies. However, existing classical network protocols cannot simply be adapted to quantum networks. Current approaches do not adequately tackle the difficult real-world issues inherent in quantum systems, such as the probabilistic nature of entanglement generation, rapid decoherence of quantum states, and the high cost of performing measurements on noisy intermediate-scale quantum (NISQ) hardware. Given a specific quantum network architecture, the underlying theoretical issues of how to optimally allocate probes for network tomography or sequentially learn the best routing paths with minimal quantum resource consumption remain critical open challenges.
Building on our recent breakthroughs in quantum best arm identification and online optimal path learning, this project aims to develop a rigorous, sequential decision-making framework tailored for the unique dynamics of quantum networks. We will utilize an information-theoretic and online optimization approach to evaluate internal quantum channels and optimize routing topologies.
Objectives:
-
Online Path Learning: Design a novel online optimization framework that continuously learns and updates the best entanglement routing paths in a dynamic quantum network, specifically minimizing the consumption of high-fidelity qubits.
-
Optimal Tomography: Theoretically formulate the optimal probe allocation strategies for quantum network tomography using multi-armed bandit theories to accurately evaluate channel capacities with minimal sample complexity.
-
Empirical Validation: Validate the designed sequential decision models and tomography algorithms by conducting systematic experiments on leading quantum network simulators (e.g., SeQUeNCe or NetSquid) against simulated NISQ noise models.
Findings So Far:
We have established foundational results for decision-making in quantum environments. Our recent work on "Online Optimal Probe Allocation" [1] introduces a sequential strategy for quantum network tomography, allowing for the characterization of internal quantum channels with minimal overhead. This builds upon our theoretical framework for "Best Arm Identification with Quantum Oracles" [2], which proves that quantum speedups can be utilized to identify optimal network parameters with significantly lower query complexity than classical approaches. Additionally, we have pioneered algorithms for "Learning Best Paths in Quantum Networks" [3]. Our experimental results on simulated multi-hop topologies demonstrate that our adaptive algorithms converge to optimal routing paths faster than fixed-protocol baselines, significantly increasing the entanglement success rate in resource-constrained environments.
Selected Publications:
[1] X. Wang, Y.-Z. J. Chen, M. Andrade, M. Hajiesmaili, J. C.S. Lui, T. He, and D. Towsley, "Online Optimal Probe Allocation for Quantum Network Tomography," International Conference on Quantum Communications, Networking, and Computing (QCNC), 2026.
[2] X. Wang, Y.-Z. J. Chen, M. Andrade, J. Allcock, M. Hajiesmaili, J. C.S. Lui, and D. Towsley, "Best Arm Identification with Quantum Oracles," The 39th Annual AAAI Conference on Artificial Intelligence (AAAI), 2025.
[3] X. Wang, M. Liu, X. Liu, Z. Li, M. Hajiesmaili, J. C.S. Lui, and D. Towsley, "Learning Best Paths in Quantum Networks," Proceedings of the IEEE Conference on Computer Communications (INFOCOM), 2025.
For further information on this research topic, please contact Prof. Xuchuang Wang.
 |