Entanglement Distribution Delay Optimization in Quantum Networks with Distillation

Mahdi Chehimi, Kenneth Goodenough, Walid Saad, Don Towsley, Tony X. Zhou

Abstract

Quantum networks (QNs) face significant challenges due to the fragility of entangled states, which suffer from decoherence and loss during transmission and storage. While entanglement distillation is a vital technique to improve fidelity, it introduces substantial distribution delays. This paper proposes a resource allocation framework that jointly optimizes entanglement distribution delay and distillation operations by incorporating the specific physical properties of nitrogen-vacancy (NV) centers in diamond. Results demonstrate a 30-50% reduction in delay and an 11% improvement in end-to-end fidelity compared to hardware-agnostic benchmarks.

Introduction

The goal of quantum networks is the reliable distribution of entangled states across distant nodes to support applications in secure communication and distributed sensing. However, quantum states are susceptible to noise and loss during transmission through optical fibers and storage in memory. Entanglement distillation addresses this by combining multiple low-quality pairs into a single high-quality pair. While effective for fidelity, distillation consumes time and hardware resources, creating a bottleneck. Most existing research focuses on maximizing rate or fidelity in isolation, often ignoring the delay introduced by distillation or the specific hardware characteristics of the qubits.

Methodology

The authors model a star-shaped quantum network centered around a Quantum Switch (QS). The framework integrates physics-based modeling of Nitrogen-Vacancy (NV) centers in diamond, allowing the switch to select between different NV center types and nuclear spin regions for memory. This selection balances gate speeds against coherence times.

System architecture of the proposed star-shaped quantum network featuring NV center hardware details.

The technical approach involves three primary components:

  1. Analytical Noise Derivation: A closed-form expression for average decoherence noise is derived using a "twirling map" to simplify non-uniform noise into a manageable depolarizing parameter.
  2. Joint Optimization: A min-max optimization problem is formulated to minimize the maximum delay across all users. Variables include entanglement generation parameters, hardware types, and distillation protocols.
  3. Simulated Annealing: Due to the NP-hard nature of the optimization, a simulated annealing metaheuristic is used to find near-optimal solutions efficiently.

Key Insights

A critical finding is the "distillation tradeoff": while necessary for distant users, distillation can be detrimental for users near the switch where noise is minimal. Furthermore, the switch must optimize the physical environment of the qubit. Controlling the isotopic purity of the diamond and selecting specific nuclear spins for memory provides essential "knobs" to balance gate noise against decoherence.

Results

Numerical simulations using parameters from experimental literature (e.g., gate fidelities of 0.97–0.999 and attenuation lengths of 0.542 km) show that the proposed framework significantly outperforms benchmarks:

  • Delay Reduction: The framework achieved a 30% reduction in delay compared to physics-agnostic models and a 50% reduction compared to models that always perform minimal distillation.
  • Fidelity Gains: Average end-to-end fidelity improved by 5% to 11% over existing frameworks.
  • Reliability: The framework was uniquely capable of meeting strict fidelity and rate requirements for all users, particularly at greater distances from the switch.

Conclusion and Future Work

This work highlights the importance of hardware-aware optimization in quantum network design. By considering the physical nuances of NV centers, the proposed framework effectively manages the distillation-delay tradeoff. However, the study is currently limited to a star-shaped topology. Future work will extend this to mesh networks with multiple hops and explore more complex, multi-round distillation protocols.

Summary updated

BibTeX
@article{chehimi2025entanglement,
  author = {Mahdi Chehimi and Kenneth Goodenough and Walid Saad and Don Towsley and Tony X. Zhou},
  title = {{Entanglement Distribution Delay Optimization in Quantum Networks with Distillation}},
  journal = {IEEE J. Sel. Areas Commun.},
  year = {2025},
  doi = {10.1109/jsac.2025.3543485},
}
Figure